Bump version: 0.27.1 → 0.28.0-beta.0

## Summary

- PR #2957 changed the permutation builder to only select `_rowid` from
the base table, but `Splitter::project()` for hash and calculated splits
replaced the selection entirely, dropping `_rowid`.
- Include `_rowid` in the column selections for hash and calculated
split projections.
- Fix a Python test that queried the permutation table for base table
columns no longer materialized.

Fixes the `test_split_hash`, `test_split_hash_with_discard`,
`test_split_calculated`, `test_shuffle_combined_with_splits`, and
`test_filter_with_splits` failures in `test_permutation.py`.

## Test plan

- [x] `cargo test -p lancedb -- permutation` (22 passed)
- [x] `pytest python/tests/test_permutation.py` (46 passed)
- [x] `npm test __test__/permutation.test.ts` (20 passed)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.23.0-beta.0"
+current_version = "0.24.1"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

.github/workflows/build_linux_wheel/action.yml

@@ -3,7 +3,7 @@ name: build-linux-wheel
 description: "Build a manylinux wheel for lance"
 inputs:
   python-minor-version:
-    description: "8, 9, 10, 11, 12"
+    description: "10, 11, 12, 13"
     required: true
   args:
     description: "--release"

.github/workflows/build_mac_wheel/action.yml

@@ -3,7 +3,7 @@ name: build_wheel
 description: "Build a lance wheel"
 inputs:
   python-minor-version:
-    description: "8, 9, 10, 11"
+    description: "10, 11, 12, 13"
     required: true
   args:
     description: "--release"

.github/workflows/build_windows_wheel/action.yml

@@ -3,7 +3,7 @@ name: build_wheel
 description: "Build a lance wheel"
 inputs:
   python-minor-version:
-    description: "8, 9, 10, 11"
+    description: "10, 11, 12, 13, 14"
     required: true
   args:
     description: "--release"

.github/workflows/codex-update-lance-dependency.yml

@@ -75,6 +75,13 @@ jobs:
           VERSION="${VERSION#v}"
           BRANCH_NAME="codex/update-lance-${VERSION//[^a-zA-Z0-9]/-}"
 
+          # Use "chore" for beta/rc versions, "feat" for stable releases
+          if [[ "${VERSION}" == *beta* ]] || [[ "${VERSION}" == *rc* ]]; then
+            COMMIT_TYPE="chore"
+          else
+            COMMIT_TYPE="feat"
+          fi
+
           cat <<EOF >/tmp/codex-prompt.txt
           You are running inside the lancedb repository on a GitHub Actions runner. Update the Lance dependency to version ${VERSION} and prepare a pull request for maintainers to review.
 
@@ -84,10 +91,10 @@ jobs:
           3. After clippy succeeds, run "cargo fmt --all" to format the workspace.
           4. Ensure the repository is clean except for intentional changes. Inspect "git status --short" and "git diff" to confirm the dependency update and any required fixes.
           5. Create and switch to a new branch named "${BRANCH_NAME}" (replace any duplicated hyphens if necessary).
-          6. Stage all relevant files with "git add -A". Commit using the message "chore: update lance dependency to v${VERSION}".
+          6. Stage all relevant files with "git add -A". Commit using the message "${COMMIT_TYPE}: update lance dependency to v${VERSION}".
           7. Push the branch to origin. If the branch already exists, force-push your changes.
           8. env "GH_TOKEN" is available, use "gh" tools for github related operations like creating pull request.
-          9. Create a pull request targeting "main" with title "chore: update lance dependency to v${VERSION}". In the body, summarize the dependency bump, clippy/fmt verification, and link the triggering tag (${TAG}).
+          9. Create a pull request targeting "main" with title "${COMMIT_TYPE}: update lance dependency to v${VERSION}". First, write the PR body to /tmp/pr-body.md using a heredoc (cat <<'EOF' > /tmp/pr-body.md). The body should summarize the dependency bump, clippy/fmt verification, and link the triggering tag (${TAG}). Then run "gh pr create --body-file /tmp/pr-body.md".
           10. After creating the PR, display the PR URL, "git status --short", and a concise summary of the commands run and their results.
 
           Constraints:

.github/workflows/docs.yml

@@ -41,7 +41,7 @@ jobs:
           sudo apt install -y protobuf-compiler libssl-dev
           rustup update && rustup default
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
           python-version: "3.10"
           cache: "pip"

.github/workflows/nodejs.yml

@@ -88,7 +88,7 @@ jobs:
         npm install -g @napi-rs/cli
     - name: Build
       run: |
-        npm ci
+        npm ci --include=optional
         npm run build:debug -- --profile ci
         npm run tsc
     - name: Setup localstack
@@ -146,7 +146,7 @@ jobs:
         npm install -g @napi-rs/cli
     - name: Build
       run: |
-        npm ci
+        npm ci --include=optional
         npm run build:debug -- --profile ci
         npm run tsc
     - name: Test

.github/workflows/pypi-publish.yml

@@ -44,12 +44,12 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v6
         with:
-          python-version: 3.8
+          python-version: "3.10"
       - uses: ./.github/workflows/build_linux_wheel
         with:
-          python-minor-version: 8
+          python-minor-version: 10
           args: "--release --strip ${{ matrix.config.extra_args }}"
           arm-build: ${{ matrix.config.platform == 'aarch64' }}
           manylinux: ${{ matrix.config.manylinux }}
@@ -74,12 +74,12 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v6
         with:
-          python-version: 3.12
+          python-version: "3.13"
       - uses: ./.github/workflows/build_mac_wheel
         with:
-          python-minor-version: 8
+          python-minor-version: 10
           args: "--release --strip --target ${{ matrix.config.target }} --features fp16kernels"
       - uses: ./.github/workflows/upload_wheel
         if: startsWith(github.ref, 'refs/tags/python-v')
@@ -95,12 +95,12 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v6
         with:
-          python-version: 3.12
+          python-version: "3.13"
       - uses: ./.github/workflows/build_windows_wheel
         with:
-          python-minor-version: 8
+          python-minor-version: 10
           args: "--release --strip"
           vcpkg_token: ${{ secrets.VCPKG_GITHUB_PACKAGES }}
       - uses: ./.github/workflows/upload_wheel

.github/workflows/python.yml

@@ -36,9 +36,9 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
-          python-version: "3.12"
+          python-version: "3.13"
       - name: Install ruff
         run: |
           pip install ruff==0.9.9
@@ -49,8 +49,8 @@ jobs:
 
   type-check:
     name: "Type Check"
-    timeout-minutes: 30
-    runs-on: "ubuntu-22.04"
+    timeout-minutes: 60
+    runs-on: ubuntu-2404-8x-x64
     defaults:
       run:
         shell: bash
@@ -61,9 +61,9 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
-          python-version: "3.12"
+          python-version: "3.13"
       - name: Install protobuf compiler
         run: |
           sudo apt update
@@ -78,7 +78,7 @@ jobs:
 
   doctest:
     name: "Doctest"
-    timeout-minutes: 30
+    timeout-minutes: 60
     runs-on: ubuntu-2404-8x-x64
     defaults:
       run:
@@ -90,9 +90,9 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
-          python-version: "3.12"
+          python-version: "3.13"
           cache: "pip"
       - name: Install protobuf
         run: |
@@ -110,7 +110,7 @@ jobs:
     timeout-minutes: 30
     strategy:
       matrix:
-        python-minor-version: ["9", "12"]
+        python-minor-version: ["10", "13"]
     runs-on: "ubuntu-24.04"
     defaults:
       run:
@@ -126,7 +126,7 @@ jobs:
           sudo apt update
           sudo apt install -y protobuf-compiler
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
           python-version: 3.${{ matrix.python-minor-version }}
       - uses: ./.github/workflows/build_linux_wheel
@@ -156,9 +156,9 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
-          python-version: "3.12"
+          python-version: "3.13"
       - uses: ./.github/workflows/build_mac_wheel
         with:
           args: --profile ci
@@ -185,9 +185,9 @@ jobs:
           fetch-depth: 0
           lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
-          python-version: "3.12"
+          python-version: "3.13"
       - uses: ./.github/workflows/build_windows_wheel
         with:
           args: --profile ci
@@ -212,9 +212,9 @@ jobs:
           sudo apt update
           sudo apt install -y protobuf-compiler
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
-          python-version: 3.9
+          python-version: "3.10"
       - name: Install lancedb
         run: |
           pip install "pydantic<2"

.github/workflows/rust.yml

@@ -48,6 +48,8 @@ jobs:
         run: cargo fmt --all -- --check
       - name: Run clippy
         run: cargo clippy --profile ci --workspace --tests --all-features -- -D warnings
+      - name: Run clippy (without remote feature)
+        run: cargo clippy --profile ci --workspace --tests -- -D warnings
 
   build-no-lock:
     runs-on: ubuntu-24.04
@@ -167,13 +169,13 @@ jobs:
       - name: Build
         run: |
           $env:VCPKG_ROOT = $env:VCPKG_INSTALLATION_ROOT
-          cargo build --profile ci --features remote --tests --locked --target ${{ matrix.target }}
+          cargo build --profile ci --features aws,remote --tests --locked --target ${{ matrix.target }}
       - name: Run tests
         # Can only run tests when target matches host
         if: ${{ matrix.target == 'x86_64-pc-windows-msvc' }}
         run: |
           $env:VCPKG_ROOT = $env:VCPKG_INSTALLATION_ROOT
-          cargo test --profile ci --features remote --locked
+          cargo test --profile ci --features aws,remote --locked
 
   msrv:
     # Check the minimum supported Rust version
@@ -181,7 +183,7 @@ jobs:
     runs-on: ubuntu-24.04
     strategy:
       matrix:
-        msrv: ["1.78.0"] # This should match up with rust-version in Cargo.toml
+        msrv: ["1.88.0"] # This should match up with rust-version in Cargo.toml
     env:
       # Need up-to-date compilers for kernels
       CC: clang-18
@@ -212,4 +214,6 @@ jobs:
           cargo update -p aws-sdk-sts --precise 1.51.0
           cargo update -p home --precise 0.5.9
       - name: cargo +${{ matrix.msrv }} check
+        env:
+          RUSTUP_TOOLCHAIN: ${{ matrix.msrv }}
         run: cargo check --profile ci --workspace --tests --benches --all-features

Cargo.lock

@@ -1,6 +1,6 @@
 # This file is automatically @generated by Cargo.
 # It is not intended for manual editing.
-version = 3
+version = 4
 
 [[package]]
 name = "adler2"
@@ -1041,6 +1041,61 @@ dependencies = [
  "tracing",
 ]
 
+[[package]]
+name = "axum"
+version = "0.7.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "edca88bc138befd0323b20752846e6587272d3b03b0343c8ea28a6f819e6e71f"
+dependencies = [
+ "async-trait",
+ "axum-core",
+ "bytes",
+ "futures-util",
+ "http 1.3.1",
+ "http-body 1.0.1",
+ "http-body-util",
+ "hyper 1.7.0",
+ "hyper-util",
+ "itoa",
+ "matchit",
+ "memchr",
+ "mime",
+ "percent-encoding",
+ "pin-project-lite",
+ "rustversion",
+ "serde",
+ "serde_json",
+ "serde_path_to_error",
+ "serde_urlencoded",
+ "sync_wrapper",
+ "tokio",
+ "tower",
+ "tower-layer",
+ "tower-service",
+ "tracing",
+]
+
+[[package]]
+name = "axum-core"
+version = "0.4.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "09f2bd6146b97ae3359fa0cc6d6b376d9539582c7b4220f041a33ec24c226199"
+dependencies = [
+ "async-trait",
+ "bytes",
+ "futures-util",
+ "http 1.3.1",
+ "http-body 1.0.1",
+ "http-body-util",
+ "mime",
+ "pin-project-lite",
+ "rustversion",
+ "sync_wrapper",
+ "tower-layer",
+ "tower-service",
+ "tracing",
+]
+
 [[package]]
 name = "backoff"
 version = "0.4.0"
@@ -3086,8 +3141,8 @@ checksum = "42703706b716c37f96a77aea830392ad231f44c9e9a67872fa5548707e11b11c"
 
 [[package]]
 name = "fsst"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-array",
  "rand 0.9.2",
@@ -3930,6 +3985,7 @@ dependencies = [
  "http 1.3.1",
  "http-body 1.0.1",
  "httparse",
+ "httpdate",
  "itoa",
  "pin-project-lite",
  "pin-utils",
@@ -4205,7 +4261,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4b0f83760fb341a774ed326568e19f5a863af4a952def8c39f9ab92fd95b88e5"
 dependencies = [
  "equivalent",
- "hashbrown 0.15.5",
+ "hashbrown 0.16.0",
  "serde",
  "serde_core",
 ]
@@ -4422,8 +4478,8 @@ dependencies = [
 
 [[package]]
 name = "lance"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4488,8 +4544,8 @@ dependencies = [
 
 [[package]]
 name = "lance-arrow"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4507,8 +4563,8 @@ dependencies = [
 
 [[package]]
 name = "lance-bitpacking"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrayref",
  "paste",
@@ -4517,8 +4573,8 @@ dependencies = [
 
 [[package]]
 name = "lance-core"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4554,8 +4610,8 @@ dependencies = [
 
 [[package]]
 name = "lance-datafusion"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4585,8 +4641,8 @@ dependencies = [
 
 [[package]]
 name = "lance-datagen"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4603,8 +4659,8 @@ dependencies = [
 
 [[package]]
 name = "lance-encoding"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -4641,8 +4697,8 @@ dependencies = [
 
 [[package]]
 name = "lance-file"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -4674,8 +4730,8 @@ dependencies = [
 
 [[package]]
 name = "lance-geo"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "datafusion",
  "geo-types",
@@ -4686,8 +4742,8 @@ dependencies = [
 
 [[package]]
 name = "lance-index"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4748,8 +4804,8 @@ dependencies = [
 
 [[package]]
 name = "lance-io"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-arith",
@@ -4789,8 +4845,8 @@ dependencies = [
 
 [[package]]
 name = "lance-linalg"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4806,8 +4862,8 @@ dependencies = [
 
 [[package]]
 name = "lance-namespace"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "async-trait",
@@ -4819,14 +4875,16 @@ dependencies = [
 
 [[package]]
 name = "lance-namespace-impls"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-ipc",
  "arrow-schema",
  "async-trait",
+ "axum",
  "bytes",
+ "chrono",
  "futures",
  "lance",
  "lance-core",
@@ -4837,17 +4895,20 @@ dependencies = [
  "object_store",
  "rand 0.9.2",
  "reqwest",
+ "serde",
  "serde_json",
  "snafu",
  "tokio",
+ "tower",
+ "tower-http 0.5.2",
  "url",
 ]
 
 [[package]]
 name = "lance-namespace-reqwest-client"
-version = "0.0.18"
+version = "0.4.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3ea349999bcda4eea53fc05d334b3775ec314761e6a706555c777d7a29b18d19"
+checksum = "a2acdba67f84190067532fce07b51a435dd390d7cdc1129a05003e5cb3274cf0"
 dependencies = [
  "reqwest",
  "serde",
@@ -4858,8 +4919,8 @@ dependencies = [
 
 [[package]]
 name = "lance-table"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow",
  "arrow-array",
@@ -4898,8 +4959,8 @@ dependencies = [
 
 [[package]]
 name = "lance-testing"
-version = "1.1.0-beta.1"
-source = "git+https://github.com/lance-format/lance.git?tag=v1.1.0-beta.1#ddea38f049e64df8b893e1c8ecca7878ea373d1e"
+version = "1.0.4"
+source = "git+https://github.com/lance-format/lance.git?tag=v1.0.4#a93eaad1f6909a843cf8aa00d5530359012a7aaa"
 dependencies = [
  "arrow-array",
  "arrow-schema",
@@ -4910,7 +4971,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb"
-version = "0.23.0-beta.0"
+version = "0.24.1"
 dependencies = [
  "ahash",
  "anyhow",
@@ -4989,7 +5050,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb-nodejs"
-version = "0.23.0-beta.0"
+version = "0.24.1"
 dependencies = [
  "arrow-array",
  "arrow-ipc",
@@ -5009,7 +5070,7 @@ dependencies = [
 
 [[package]]
 name = "lancedb-python"
-version = "0.26.0-beta.0"
+version = "0.27.1"
 dependencies = [
  "arrow",
  "async-trait",
@@ -5277,6 +5338,12 @@ dependencies = [
  "regex-automata",
 ]
 
+[[package]]
+name = "matchit"
+version = "0.7.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0e7465ac9959cc2b1404e8e2367b43684a6d13790fe23056cc8c6c5a6b7bcb94"
+
 [[package]]
 name = "matrixmultiply"
 version = "0.3.10"
@@ -6659,8 +6726,8 @@ version = "0.13.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "be769465445e8c1474e9c5dac2018218498557af32d9ed057325ec9a41ae81bf"
 dependencies = [
- "heck 0.4.1",
- "itertools 0.12.1",
+ "heck 0.5.0",
+ "itertools 0.14.0",
  "log",
  "multimap",
  "once_cell",
@@ -6680,7 +6747,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8a56d757972c98b346a9b766e3f02746cde6dd1cd1d1d563472929fdd74bec4d"
 dependencies = [
  "anyhow",
- "itertools 0.12.1",
+ "itertools 0.14.0",
  "proc-macro2",
  "quote",
  "syn 2.0.106",
@@ -7265,7 +7332,7 @@ dependencies = [
  "tokio-rustls 0.26.4",
  "tokio-util",
  "tower",
- "tower-http",
+ "tower-http 0.6.6",
  "tower-service",
  "url",
  "wasm-bindgen",
@@ -7784,6 +7851,17 @@ dependencies = [
  "serde_core",
 ]
 
+[[package]]
+name = "serde_path_to_error"
+version = "0.1.20"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "10a9ff822e371bb5403e391ecd83e182e0e77ba7f6fe0160b795797109d1b457"
+dependencies = [
+ "itoa",
+ "serde",
+ "serde_core",
+]
+
 [[package]]
 name = "serde_plain"
 version = "1.0.2"
@@ -7999,7 +8077,7 @@ version = "0.8.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c1c97747dbf44bb1ca44a561ece23508e99cb592e862f22222dcf42f51d1e451"
 dependencies = [
- "heck 0.4.1",
+ "heck 0.5.0",
  "proc-macro2",
  "quote",
  "syn 2.0.106",
@@ -8819,6 +8897,24 @@ dependencies = [
  "tokio",
  "tower-layer",
  "tower-service",
+ "tracing",
+]
+
+[[package]]
+name = "tower-http"
+version = "0.5.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1e9cd434a998747dd2c4276bc96ee2e0c7a2eadf3cae88e52be55a05fa9053f5"
+dependencies = [
+ "bitflags 2.9.4",
+ "bytes",
+ "http 1.3.1",
+ "http-body 1.0.1",
+ "http-body-util",
+ "pin-project-lite",
+ "tower-layer",
+ "tower-service",
+ "tracing",
 ]
 
 [[package]]
@@ -8857,6 +8953,7 @@ version = "0.1.41"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "784e0ac535deb450455cbfa28a6f0df145ea1bb7ae51b821cf5e7927fdcfbdd0"
 dependencies = [
+ "log",
  "pin-project-lite",
  "tracing-attributes",
  "tracing-core",

Cargo.toml

@@ -12,23 +12,23 @@ repository = "https://github.com/lancedb/lancedb"
 description = "Serverless, low-latency vector database for AI applications"
 keywords = ["lancedb", "lance", "database", "vector", "search"]
 categories = ["database-implementations"]
-rust-version = "1.78.0"
+rust-version = "1.88.0"
 
 [workspace.dependencies]
-lance = { "version" = "=1.1.0-beta.1", default-features = false, "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-core = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-datagen = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-file = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-io = { "version" = "=1.1.0-beta.1", default-features = false, "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-index = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-linalg = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-namespace = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-namespace-impls = { "version" = "=1.1.0-beta.1", default-features = false, "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-table = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-testing = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-datafusion = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-encoding = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
-lance-arrow = { "version" = "=1.1.0-beta.1", "tag" = "v1.1.0-beta.1", "git" = "https://github.com/lance-format/lance.git" }
+lance = { "version" = "=1.0.4", default-features = false, "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-core = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-datagen = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-file = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-io = { "version" = "=1.0.4", default-features = false, "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-index = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-linalg = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-namespace = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-namespace-impls = { "version" = "=1.0.4", default-features = false, "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-table = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-testing = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-datafusion = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-encoding = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
+lance-arrow = { "version" = "=1.0.4", "tag" = "v1.0.4", "git" = "https://github.com/lance-format/lance.git" }
 ahash = "0.8"
 # Note that this one does not include pyarrow
 arrow = { version = "56.2", optional = false }

README.md

@@ -66,7 +66,7 @@ Follow the [Quickstart](https://lancedb.com/docs/quickstart/) doc to set up Lanc
 | Python SDK | https://lancedb.github.io/lancedb/python/python/ |
 | Typescript SDK | https://lancedb.github.io/lancedb/js/globals/ |
 | Rust SDK | https://docs.rs/lancedb/latest/lancedb/index.html |
-| REST API | https://docs.lancedb.com/api-reference/introduction |
+| REST API | https://docs.lancedb.com/api-reference/rest |
 
 ## **Join Us and Contribute**
 

ci/run_with_test_connection.sh

@@ -16,7 +16,7 @@ check_command_exists() {
 }
 
 if [[ ! -e ./lancedb ]]; then
-    if [[ -v SOPHON_READ_TOKEN ]]; then
+    if [[ x${SOPHON_READ_TOKEN} != "x" ]]; then
         INPUT="lancedb-linux-x64"
         gh release \
             --repo lancedb/lancedb \

ci/set_lance_version.py

@@ -229,6 +229,29 @@ def set_local_version():
     update_cargo_toml(line_updater)
 
 
+def update_lockfiles(version: str, fallback_to_git: bool = False):
+    """
+    Update Cargo metadata and optionally fall back to using the git tag if the
+    requested crates.io version is unavailable.
+    """
+    try:
+        print("Updating lockfiles...", file=sys.stderr, end="")
+        run_command("cargo metadata > /dev/null")
+        print(" done.", file=sys.stderr)
+    except Exception as e:
+        if fallback_to_git and "failed to select a version" in str(e):
+            print(
+                f" failed for crates.io v{version}, retrying with git tag...",
+                file=sys.stderr,
+            )
+            set_preview_version(version)
+            print("Updating lockfiles...", file=sys.stderr, end="")
+            run_command("cargo metadata > /dev/null")
+            print(" done.", file=sys.stderr)
+        else:
+            raise
+
+
 parser = argparse.ArgumentParser(description="Set the version of the Lance package.")
 parser.add_argument(
     "version",
@@ -244,6 +267,7 @@ if args.version == "stable":
         file=sys.stderr,
     )
     set_stable_version(latest_stable_version)
+    update_lockfiles(latest_stable_version)
 elif args.version == "preview":
     latest_preview_version = get_latest_preview_version()
     print(
@@ -251,8 +275,10 @@ elif args.version == "preview":
         file=sys.stderr,
     )
     set_preview_version(latest_preview_version)
+    update_lockfiles(latest_preview_version)
 elif args.version == "local":
     set_local_version()
+    update_lockfiles("local")
 else:
     # Parse the version number.
     version = args.version
@@ -262,9 +288,7 @@ else:
 
     if "beta" in version:
         set_preview_version(version)
+        update_lockfiles(version)
     else:
         set_stable_version(version)
-
-print("Updating lockfiles...", file=sys.stderr, end="")
-run_command("cargo metadata > /dev/null")
-print(" done.", file=sys.stderr)
+        update_lockfiles(version, fallback_to_git=True)

docs/mkdocs.yml

@@ -11,7 +11,7 @@ watch:
 theme:
   name: "material"
   logo: assets/logo.png
-  favicon: assets/logo.png
+  favicon: assets/favicon.ico
   palette:
     # Palette toggle for light mode
     - scheme: lancedb
@@ -32,8 +32,6 @@ theme:
     - content.tooltips
     - toc.follow
     - navigation.top
-    - navigation.tabs
-    - navigation.tabs.sticky
     - navigation.footer
     - navigation.tracking
     - navigation.instant
@@ -115,12 +113,13 @@ markdown_extensions:
       emoji_index: !!python/name:material.extensions.emoji.twemoji
       emoji_generator: !!python/name:material.extensions.emoji.to_svg
   - markdown.extensions.toc:
-      baselevel: 1
-      permalink: ""
+      toc_depth: 3
+      permalink: true
+      permalink_title: Anchor link to this section
 
 nav:
-  - API reference:
-      - Overview: index.md
+  - Documentation:
+      - SDK Reference: index.md
       - Python: python/python.md
       - Javascript/TypeScript: js/globals.md
       - Java: java/java.md

docs/src/embeddings/available_embedding_models/multimodal_embedding_functions/voyageai_multimodal_embedding.md

@@ -0,0 +1,111 @@
+# VoyageAI Embeddings : Multimodal
+
+VoyageAI embeddings can also be used to embed both text and image data, only some of the models support image data and you can check the list
+under [https://docs.voyageai.com/docs/multimodal-embeddings](https://docs.voyageai.com/docs/multimodal-embeddings)
+
+Supported multimodal models:
+
+- `voyage-multimodal-3` - 1024 dimensions (text + images)
+- `voyage-multimodal-3.5` - Flexible dimensions (256, 512, 1024 default, 2048). Supports text, images, and video.
+
+### Video Support (voyage-multimodal-3.5)
+
+The `voyage-multimodal-3.5` model supports video input through:
+- Video URLs (`.mp4`, `.webm`, `.mov`, `.avi`, `.mkv`, `.m4v`, `.gif`)
+- Video file paths
+
+Constraints: Max 20MB video size.
+
+Supported parameters (to be passed in `create` method) are:
+
+| Parameter | Type | Default Value           | Description                               |
+|---|---|-------------------------|-------------------------------------------|
+| `name` | `str` | `"voyage-multimodal-3"` | The model ID of the VoyageAI model to use |
+| `output_dimension` | `int` | `None` | Output dimension for voyage-multimodal-3.5. Valid: 256, 512, 1024, 2048 |
+
+Usage Example:
+
+```python
+import base64
+import os
+from io import BytesIO
+
+import requests
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
+import pandas as pd
+
+os.environ['VOYAGE_API_KEY'] = 'YOUR_VOYAGE_API_KEY'
+
+db = lancedb.connect(".lancedb")
+func = get_registry().get("voyageai").create(name="voyage-multimodal-3")
+
+
+def image_to_base64(image_bytes: bytes):
+    buffered = BytesIO(image_bytes)
+    img_str = base64.b64encode(buffered.getvalue())
+    return img_str.decode("utf-8")
+
+
+class Images(LanceModel):
+    label: str
+    image_uri: str = func.SourceField()  # image uri as the source
+    image_bytes: str = func.SourceField()  # image bytes base64 encoded as the source
+    vector: Vector(func.ndims()) = func.VectorField()  # vector column
+    vec_from_bytes: Vector(func.ndims()) = func.VectorField()  # Another vector column
+
+
+if "images" in db.table_names():
+    db.drop_table("images")
+table = db.create_table("images", schema=Images)
+labels = ["cat", "cat", "dog", "dog", "horse", "horse"]
+uris = [
+    "http://farm1.staticflickr.com/53/167798175_7c7845bbbd_z.jpg",
+    "http://farm1.staticflickr.com/134/332220238_da527d8140_z.jpg",
+    "http://farm9.staticflickr.com/8387/8602747737_2e5c2a45d4_z.jpg",
+    "http://farm5.staticflickr.com/4092/5017326486_1f46057f5f_z.jpg",
+    "http://farm9.staticflickr.com/8216/8434969557_d37882c42d_z.jpg",
+    "http://farm6.staticflickr.com/5142/5835678453_4f3a4edb45_z.jpg",
+]
+# get each uri as bytes
+images_bytes = [image_to_base64(requests.get(uri).content) for uri in uris]
+table.add(
+    pd.DataFrame({"label": labels, "image_uri": uris, "image_bytes": images_bytes})
+)
+```
+Now we can search using text from both the default vector column and the custom vector column
+```python
+
+# text search
+actual = table.search("man's best friend", "vec_from_bytes").limit(1).to_pydantic(Images)[0]
+print(actual.label) # prints "dog"
+
+frombytes = (
+    table.search("man's best friend", vector_column_name="vec_from_bytes")
+    .limit(1)
+    .to_pydantic(Images)[0]
+)
+print(frombytes.label)
+
+```
+
+Because we're using a multi-modal embedding function, we can also search using images
+
+```python
+# image search
+query_image_uri = "http://farm1.staticflickr.com/200/467715466_ed4a31801f_z.jpg"
+image_bytes = requests.get(query_image_uri).content
+query_image = Image.open(BytesIO(image_bytes))
+actual = table.search(query_image, "vec_from_bytes").limit(1).to_pydantic(Images)[0]
+print(actual.label == "dog")
+
+# image search using a custom vector column
+other = (
+    table.search(query_image, vector_column_name="vec_from_bytes")
+    .limit(1)
+    .to_pydantic(Images)[0]
+)
+print(actual.label)
+
+```

docs/src/embeddings/available_embedding_models/text_embedding_functions/voyageai_embedding.md

@@ -0,0 +1,62 @@
+# VoyageAI Embeddings
+
+Voyage AI provides cutting-edge embedding and rerankers.
+
+
+Using voyageai API requires voyageai package, which can be installed using `pip install voyageai`. Voyage AI embeddings are used to generate embeddings for text data. The embeddings can be used for various tasks like semantic search, clustering, and classification.
+You also need to set the `VOYAGE_API_KEY` environment variable to use the VoyageAI API.
+
+Supported models are:
+
+**Voyage-4 Series (Latest)**
+
+- voyage-4 (1024 dims, general-purpose and multilingual retrieval, 320K batch tokens)
+- voyage-4-lite (1024 dims, optimized for latency and cost, 1M batch tokens)
+- voyage-4-large (1024 dims, best retrieval quality, 120K batch tokens)
+
+**Voyage-3 Series**
+
+- voyage-3
+- voyage-3-lite
+
+**Domain-Specific Models**
+
+- voyage-finance-2
+- voyage-multilingual-2
+- voyage-law-2
+- voyage-code-2
+
+
+Supported parameters (to be passed in `create` method) are:
+
+| Parameter | Type | Default Value | Description |
+|---|---|--------|---------|
+| `name` | `str` | `None` | The model ID of the model to use. Supported base models for Text Embeddings: voyage-4, voyage-4-lite, voyage-4-large, voyage-3, voyage-3-lite, voyage-finance-2, voyage-multilingual-2, voyage-law-2, voyage-code-2 |
+| `input_type` | `str` | `None` | Type of the input text. Default to None. Other options: query, document. |
+| `truncation` | `bool` | `True` | Whether to truncate the input texts to fit within the context length. |
+
+
+Usage Example:
+    
+```python
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import EmbeddingFunctionRegistry
+
+    voyageai = EmbeddingFunctionRegistry
+        .get_instance()
+        .get("voyageai")
+        .create(name="voyage-3")
+
+    class TextModel(LanceModel):
+        text: str = voyageai.SourceField()
+        vector: Vector(voyageai.ndims()) =  voyageai.VectorField()
+
+    data = [ { "text": "hello world" },
+            { "text": "goodbye world" }]
+
+    db = lancedb.connect("~/.lancedb")
+    tbl = db.create_table("test", schema=TextModel, mode="overwrite")
+
+    tbl.add(data)
+```

docs/src/index.md

@@ -1,8 +1,12 @@
-# API Reference
+# SDK Reference
 
-This page contains the API reference for the SDKs supported by the LanceDB team.
+This site contains the API reference for the client SDKs supported by [LanceDB](https://lancedb.com).
 
 - [Python](python/python.md)
 - [JavaScript/TypeScript](js/globals.md)
 - [Java](java/java.md)
-- [Rust](https://docs.rs/lancedb/latest/lancedb/index.html)
+- [Rust](https://docs.rs/lancedb/latest/lancedb/index.html)
+
+!!! info "LanceDB Documentation"
+
+    If you're looking for the full documentation of LanceDB, visit [docs.lancedb.com](https://docs.lancedb.com).

docs/src/java/java.md

@@ -14,7 +14,7 @@ Add the following dependency to your `pom.xml`:
 <dependency>
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-core</artifactId>
-    <version>0.23.0-beta.0</version>
+    <version>0.24.1</version>
 </dependency>
 ```
 

docs/src/styles/extra.css

@@ -85,17 +85,26 @@
 
 /* Header gradient (only header area) */
 .md-header {
-  background: linear-gradient(90deg, #3B2E58 0%, #F0B7C1 45%, #E55A2B 100%);
+  background: linear-gradient(90deg, #e4d8f8 0%, #F0B7C1 45%, #E55A2B 100%);
   box-shadow: inset 0 1px 0 rgba(255,255,255,0.08), 0 1px 0 rgba(0,0,0,0.08);
 }
 
+/* Improve brand title contrast on the lavender side */
+.md-header__title,
+.md-header__topic,
+.md-header__title .md-ellipsis,
+.md-header__topic .md-ellipsis {
+  color: #2b1b3a;
+  text-shadow: 0 1px 0 rgba(255, 255, 255, 0.25);
+}
+
 /* Same colors as header for tabs (that hold the text) */
 .md-tabs {
-  background: linear-gradient(90deg, #3B2E58 0%, #F0B7C1 45%, #E55A2B 100%);
+  background: linear-gradient(90deg, #e4d8f8 0%, #F0B7C1 45%, #E55A2B 100%);
 }
 
 /* Dark scheme variant */
 [data-md-color-scheme="slate"] .md-header,
 [data-md-color-scheme="slate"] .md-tabs {
-  background: linear-gradient(90deg, #3B2E58 0%, #F0B7C1 45%, #E55A2B 100%);
+  background: linear-gradient(90deg, #e4d8f8 0%, #F0B7C1 45%, #E55A2B 100%);
 }

java/lancedb-core/pom.xml

@@ -8,7 +8,7 @@
     <parent>
       <groupId>com.lancedb</groupId>
       <artifactId>lancedb-parent</artifactId>
-      <version>0.23.0-beta.0</version>
+      <version>0.24.1-final.0</version>
       <relativePath>../pom.xml</relativePath>
     </parent>
 

java/pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>com.lancedb</groupId>
     <artifactId>lancedb-parent</artifactId>
-    <version>0.23.0-beta.0</version>
+    <version>0.24.1-final.0</version>
     <packaging>pom</packaging>
     <name>${project.artifactId}</name>
     <description>LanceDB Java SDK Parent POM</description>
@@ -28,7 +28,7 @@
     <properties>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
         <arrow.version>15.0.0</arrow.version>
-        <lance-core.version>1.0.0-rc.2</lance-core.version>
+        <lance-core.version>1.0.4</lance-core.version>
         <spotless.skip>false</spotless.skip>
         <spotless.version>2.30.0</spotless.version>
         <spotless.java.googlejavaformat.version>1.7</spotless.java.googlejavaformat.version>

nodejs/Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "lancedb-nodejs"
 edition.workspace = true
-version = "0.23.0-beta.0"
+version = "0.24.1"
 license.workspace = true
 description.workspace = true
 repository.workspace = true
@@ -36,6 +36,6 @@ aws-lc-rs = "=1.13.0"
 napi-build = "2.1"
 
 [features]
-default = ["remote", "lancedb/default"]
+default = ["remote", "lancedb/aws", "lancedb/gcs", "lancedb/azure", "lancedb/dynamodb", "lancedb/oss", "lancedb/huggingface"]
 fp16kernels = ["lancedb/fp16kernels"]
 remote = ["lancedb/remote"]

nodejs/npm/darwin-arm64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-arm64",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["darwin"],
 	"cpu": ["arm64"],
 	"main": "lancedb.darwin-arm64.node",

nodejs/npm/darwin-x64/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-darwin-x64",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["darwin"],
 	"cpu": ["x64"],
 	"main": "lancedb.darwin-x64.node",

nodejs/npm/linux-arm64-gnu/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-gnu",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-gnu.node",

nodejs/npm/linux-arm64-musl/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-arm64-musl",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["linux"],
 	"cpu": ["arm64"],
 	"main": "lancedb.linux-arm64-musl.node",

nodejs/npm/linux-x64-gnu/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-gnu",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-gnu.node",

nodejs/npm/linux-x64-musl/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-linux-x64-musl",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["linux"],
 	"cpu": ["x64"],
 	"main": "lancedb.linux-x64-musl.node",

nodejs/npm/win32-arm64-msvc/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lancedb/lancedb-win32-arm64-msvc",
-  "version": "0.23.0-beta.0",
+  "version": "0.24.1",
   "os": [
     "win32"
   ],

nodejs/npm/win32-x64-msvc/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@lancedb/lancedb-win32-x64-msvc",
-	"version": "0.23.0-beta.0",
+	"version": "0.24.1",
 	"os": ["win32"],
 	"cpu": ["x64"],
 	"main": "lancedb.win32-x64-msvc.node",

nodejs/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@lancedb/lancedb",
-  "version": "0.23.0-beta.0",
+  "version": "0.24.1",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "@lancedb/lancedb",
-      "version": "0.23.0-beta.0",
+      "version": "0.24.1",
       "cpu": [
         "x64",
         "arm64"

nodejs/package.json

@@ -11,7 +11,7 @@
     "ann"
   ],
   "private": false,
-  "version": "0.23.0-beta.0",
+  "version": "0.24.1",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",

python/.bumpversion.toml

@@ -1,5 +1,5 @@
 [tool.bumpversion]
-current_version = "0.26.0-beta.1"
+current_version = "0.28.0-beta.0"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

python/CONTRIBUTING.md

@@ -16,7 +16,7 @@ The Python package is a wrapper around the Rust library, `lancedb`. We use
 
 To set up your development environment, you will need to install the following:
 
-1. Python 3.9 or later
+1. Python 3.10 or later
 2. Cargo (Rust's package manager). Use [rustup](https://rustup.rs/) to install.
 3. [protoc](https://grpc.io/docs/protoc-installation/) (Protocol Buffers compiler)
 

python/Cargo.toml

@@ -1,13 +1,13 @@
 [package]
 name = "lancedb-python"
-version = "0.26.0-beta.1"
+version = "0.28.0-beta.0"
 edition.workspace = true
 description = "Python bindings for LanceDB"
 license.workspace = true
 repository.workspace = true
 keywords.workspace = true
 categories.workspace = true
-rust-version = "1.75.0"
+rust-version = "1.88.0"
 
 [lib]
 name = "_lancedb"
@@ -21,7 +21,7 @@ lance-core.workspace = true
 lance-namespace.workspace = true
 lance-io.workspace = true
 env_logger.workspace = true
-pyo3 = { version = "0.25", features = ["extension-module", "abi3-py39"] }
+pyo3 = { version = "0.25", features = ["extension-module", "abi3-py310"] }
 pyo3-async-runtimes = { version = "0.25", features = [
     "attributes",
     "tokio-runtime",
@@ -34,10 +34,10 @@ tokio = { version = "1.40", features = ["sync"] }
 [build-dependencies]
 pyo3-build-config = { version = "0.25", features = [
     "extension-module",
-    "abi3-py39",
+    "abi3-py310",
 ] }
 
 [features]
-default = ["remote", "lancedb/default"]
+default = ["remote",  "lancedb/aws", "lancedb/gcs", "lancedb/azure", "lancedb/dynamodb", "lancedb/oss", "lancedb/huggingface"]
 fp16kernels = ["lancedb/fp16kernels"]
 remote = ["lancedb/remote"]

python/pyproject.toml

@@ -10,13 +10,13 @@ dependencies = [
     "pyarrow>=16",
     "pydantic>=1.10",
     "tqdm>=4.27.0",
-    "lance-namespace>=0.2.1"
+    "lance-namespace>=0.3.2"
 ]
 description = "lancedb"
 authors = [{ name = "LanceDB Devs", email = "dev@lancedb.com" }]
 license = { file = "LICENSE" }
 readme = "README.md"
-requires-python = ">=3.9"
+requires-python = ">=3.10"
 keywords = [
     "data-format",
     "data-science",
@@ -33,10 +33,10 @@ classifiers = [
     "Programming Language :: Python",
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3 :: Only",
-    "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
     "Topic :: Scientific/Engineering",
 ]
 
@@ -137,4 +137,4 @@ include = [
     "python/lancedb/_lancedb.pyi",
 ]
 exclude = ["python/tests/"]
-pythonVersion = "3.12"
+pythonVersion = "3.13"

python/python/lancedb/__init__.py

@@ -13,6 +13,7 @@ __version__ = importlib.metadata.version("lancedb")
 
 from ._lancedb import connect as lancedb_connect
 from .common import URI, sanitize_uri
+from urllib.parse import urlparse
 from .db import AsyncConnection, DBConnection, LanceDBConnection
 from .io import StorageOptionsProvider
 from .remote import ClientConfig
@@ -28,6 +29,39 @@ from .namespace import (
 )
 
 
+def _check_s3_bucket_with_dots(
+    uri: str, storage_options: Optional[Dict[str, str]]
+) -> None:
+    """
+    Check if an S3 URI has a bucket name containing dots and warn if no region
+    is specified. S3 buckets with dots cannot use virtual-hosted-style URLs,
+    which breaks automatic region detection.
+
+    See: https://github.com/lancedb/lancedb/issues/1898
+    """
+    if not isinstance(uri, str) or not uri.startswith("s3://"):
+        return
+
+    parsed = urlparse(uri)
+    bucket = parsed.netloc
+
+    if "." not in bucket:
+        return
+
+    # Check if region is provided in storage_options
+    region_keys = {"region", "aws_region"}
+    has_region = storage_options and any(k in storage_options for k in region_keys)
+
+    if not has_region:
+        raise ValueError(
+            f"S3 bucket name '{bucket}' contains dots, which prevents automatic "
+            f"region detection. Please specify the region explicitly via "
+            f"storage_options={{'region': '<your-region>'}} or "
+            f"storage_options={{'aws_region': '<your-region>'}}. "
+            f"See https://github.com/lancedb/lancedb/issues/1898 for details."
+        )
+
+
 def connect(
     uri: URI,
     *,
@@ -121,9 +155,11 @@ def connect(
             storage_options=storage_options,
             **kwargs,
         )
+    _check_s3_bucket_with_dots(str(uri), storage_options)
 
     if kwargs:
         raise ValueError(f"Unknown keyword arguments: {kwargs}")
+
     return LanceDBConnection(
         uri,
         read_consistency_interval=read_consistency_interval,
@@ -211,6 +247,8 @@ async def connect_async(
     if isinstance(client_config, dict):
         client_config = ClientConfig(**client_config)
 
+    _check_s3_bucket_with_dots(str(uri), storage_options)
+
     return AsyncConnection(
         await lancedb_connect(
             sanitize_uri(uri),

python/python/lancedb/_lancedb.pyi

@@ -179,6 +179,7 @@ class Table:
         cleanup_since_ms: Optional[int] = None,
         delete_unverified: Optional[bool] = None,
     ) -> OptimizeStats: ...
+    async def uri(self) -> str: ...
     @property
     def tags(self) -> Tags: ...
     def query(self) -> Query: ...

python/python/lancedb/background_loop.py

@@ -22,7 +22,12 @@ class BackgroundEventLoop:
         self.thread.start()
 
     def run(self, future):
-        return asyncio.run_coroutine_threadsafe(future, self.loop).result()
+        concurrent_future = asyncio.run_coroutine_threadsafe(future, self.loop)
+        try:
+            return concurrent_future.result()
+        except BaseException:
+            concurrent_future.cancel()
+            raise
 
 
 LOOP = BackgroundEventLoop()

python/python/lancedb/db.py

@@ -210,10 +210,8 @@ class DBConnection(EnforceOverrides):
         page_token: str, optional
             The token to use for pagination. If not present, start from the beginning.
             Typically, this token is last table name from the previous page.
-            Only supported by LanceDb Cloud.
         limit: int, default 10
             The size of the page to return.
-            Only supported by LanceDb Cloud.
 
         Returns
         -------

python/python/lancedb/embeddings/colpali.py

@@ -275,7 +275,7 @@ class ColPaliEmbeddings(EmbeddingFunction):
         """
         Convert image inputs to PIL Images.
         """
-        PIL = attempt_import_or_raise("PIL", "pillow")
+        PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
         requests = attempt_import_or_raise("requests", "requests")
         images = self.sanitize_input(images)
         pil_images = []
@@ -285,12 +285,12 @@ class ColPaliEmbeddings(EmbeddingFunction):
                     if image.startswith(("http://", "https://")):
                         response = requests.get(image, timeout=10)
                         response.raise_for_status()
-                        pil_images.append(PIL.Image.open(io.BytesIO(response.content)))
+                        pil_images.append(PIL_Image.open(io.BytesIO(response.content)))
                     else:
-                        with PIL.Image.open(image) as im:
+                        with PIL_Image.open(image) as im:
                             pil_images.append(im.copy())
                 elif isinstance(image, bytes):
-                    pil_images.append(PIL.Image.open(io.BytesIO(image)))
+                    pil_images.append(PIL_Image.open(io.BytesIO(image)))
                 else:
                     # Assume it's a PIL Image; will raise if invalid
                     pil_images.append(image)

python/python/lancedb/embeddings/jinaai.py

@@ -77,8 +77,8 @@ class JinaEmbeddings(EmbeddingFunction):
             if isinstance(inputs, list):
                 inputs = inputs
             else:
-                PIL = attempt_import_or_raise("PIL", "pillow")
-                if isinstance(inputs, PIL.Image.Image):
+                PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
+                if isinstance(inputs, PIL_Image.Image):
                     inputs = [inputs]
         return inputs
 
@@ -89,13 +89,13 @@ class JinaEmbeddings(EmbeddingFunction):
         elif isinstance(image, (str, Path)):
             parsed = urlparse.urlparse(image)
             # TODO handle drive letter on windows.
-            PIL = attempt_import_or_raise("PIL", "pillow")
+            PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
             if parsed.scheme == "file":
-                pil_image = PIL.Image.open(parsed.path)
+                pil_image = PIL_Image.open(parsed.path)
             elif parsed.scheme == "":
-                pil_image = PIL.Image.open(image if os.name == "nt" else parsed.path)
+                pil_image = PIL_Image.open(image if os.name == "nt" else parsed.path)
             elif parsed.scheme.startswith("http"):
-                pil_image = PIL.Image.open(io.BytesIO(url_retrieve(image)))
+                pil_image = PIL_Image.open(io.BytesIO(url_retrieve(image)))
             else:
                 raise NotImplementedError("Only local and http(s) urls are supported")
             buffered = io.BytesIO()
@@ -103,9 +103,9 @@ class JinaEmbeddings(EmbeddingFunction):
             image_bytes = buffered.getvalue()
             image_dict = {"image": base64.b64encode(image_bytes).decode("utf-8")}
         else:
-            PIL = attempt_import_or_raise("PIL", "pillow")
+            PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
 
-            if isinstance(image, PIL.Image.Image):
+            if isinstance(image, PIL_Image.Image):
                 buffered = io.BytesIO()
                 image.save(buffered, format="PNG")
                 image_bytes = buffered.getvalue()
@@ -136,9 +136,9 @@ class JinaEmbeddings(EmbeddingFunction):
         elif isinstance(query, (Path, bytes)):
             return [self.generate_image_embedding(query)]
         else:
-            PIL = attempt_import_or_raise("PIL", "pillow")
+            PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
 
-            if isinstance(query, PIL.Image.Image):
+            if isinstance(query, PIL_Image.Image):
                 return [self.generate_image_embedding(query)]
             else:
                 raise TypeError(

python/python/lancedb/embeddings/open_clip.py

@@ -71,8 +71,8 @@ class OpenClipEmbeddings(EmbeddingFunction):
         if isinstance(query, str):
             return [self.generate_text_embeddings(query)]
         else:
-            PIL = attempt_import_or_raise("PIL", "pillow")
-            if isinstance(query, PIL.Image.Image):
+            PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
+            if isinstance(query, PIL_Image.Image):
                 return [self.generate_image_embedding(query)]
             else:
                 raise TypeError("OpenClip supports str or PIL Image as query")
@@ -145,20 +145,20 @@ class OpenClipEmbeddings(EmbeddingFunction):
             return self._encode_and_normalize_image(image)
 
     def _to_pil(self, image: Union[str, bytes]):
-        PIL = attempt_import_or_raise("PIL", "pillow")
+        PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
         if isinstance(image, bytes):
-            return PIL.Image.open(io.BytesIO(image))
-        if isinstance(image, PIL.Image.Image):
+            return PIL_Image.open(io.BytesIO(image))
+        if isinstance(image, PIL_Image.Image):
             return image
         elif isinstance(image, str):
             parsed = urlparse.urlparse(image)
             # TODO handle drive letter on windows.
             if parsed.scheme == "file":
-                return PIL.Image.open(parsed.path)
+                return PIL_Image.open(parsed.path)
             elif parsed.scheme == "":
-                return PIL.Image.open(image if os.name == "nt" else parsed.path)
+                return PIL_Image.open(image if os.name == "nt" else parsed.path)
             elif parsed.scheme.startswith("http"):
-                return PIL.Image.open(io.BytesIO(url_retrieve(image)))
+                return PIL_Image.open(io.BytesIO(url_retrieve(image)))
             else:
                 raise NotImplementedError("Only local and http(s) urls are supported")
 

python/python/lancedb/embeddings/siglip.py

@@ -56,8 +56,8 @@ class SigLipEmbeddings(EmbeddingFunction):
         if isinstance(query, str):
             return [self.generate_text_embeddings(query)]
         else:
-            PIL = attempt_import_or_raise("PIL", "pillow")
-            if isinstance(query, PIL.Image.Image):
+            PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
+            if isinstance(query, PIL_Image.Image):
                 return [self.generate_image_embedding(query)]
             else:
                 raise TypeError("SigLIP supports str or PIL Image as query")
@@ -127,21 +127,21 @@ class SigLipEmbeddings(EmbeddingFunction):
             return image_features.cpu().detach().numpy().squeeze()
 
     def _to_pil(self, image: Union[str, bytes, "PIL.Image.Image"]):
-        PIL = attempt_import_or_raise("PIL", "pillow")
-        if isinstance(image, PIL.Image.Image):
+        PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
+        if isinstance(image, PIL_Image.Image):
             return image.convert("RGB") if image.mode != "RGB" else image
         elif isinstance(image, bytes):
-            return PIL.Image.open(io.BytesIO(image)).convert("RGB")
+            return PIL_Image.open(io.BytesIO(image)).convert("RGB")
         elif isinstance(image, str):
             parsed = urlparse.urlparse(image)
             if parsed.scheme == "file":
-                return PIL.Image.open(parsed.path).convert("RGB")
+                return PIL_Image.open(parsed.path).convert("RGB")
             elif parsed.scheme == "":
                 path = image if os.name == "nt" else parsed.path
-                return PIL.Image.open(path).convert("RGB")
+                return PIL_Image.open(path).convert("RGB")
             elif parsed.scheme.startswith("http"):
                 image_bytes = url_retrieve(image)
-                return PIL.Image.open(io.BytesIO(image_bytes)).convert("RGB")
+                return PIL_Image.open(io.BytesIO(image_bytes)).convert("RGB")
             else:
                 raise NotImplementedError("Only local and http(s) urls are supported")
         else:

python/python/lancedb/embeddings/voyageai.py

@@ -2,7 +2,7 @@
 # SPDX-FileCopyrightText: Copyright The LanceDB Authors
 import base64
 import os
-from typing import ClassVar, TYPE_CHECKING, List, Union, Any, Generator
+from typing import ClassVar, TYPE_CHECKING, List, Union, Any, Generator, Optional
 
 from pathlib import Path
 from urllib.parse import urlparse
@@ -21,6 +21,9 @@ if TYPE_CHECKING:
 
 # Token limits for different VoyageAI models
 VOYAGE_TOTAL_TOKEN_LIMITS = {
+    "voyage-4": 320_000,
+    "voyage-4-lite": 1_000_000,
+    "voyage-4-large": 120_000,
     "voyage-context-3": 32_000,
     "voyage-3.5-lite": 1_000_000,
     "voyage-3.5": 320_000,
@@ -45,14 +48,32 @@ def is_valid_url(text):
         return False
 
 
+VIDEO_EXTENSIONS = {".mp4", ".webm", ".mov", ".avi", ".mkv", ".m4v", ".gif"}
+
+
+def is_video_url(url: str) -> bool:
+    """Check if URL points to a video file based on extension."""
+    parsed = urlparse(url)
+    path = parsed.path.lower()
+    return any(path.endswith(ext) for ext in VIDEO_EXTENSIONS)
+
+
+def is_video_path(path: Path) -> bool:
+    """Check if file path is a video file based on extension."""
+    return path.suffix.lower() in VIDEO_EXTENSIONS
+
+
 def transform_input(input_data: Union[str, bytes, Path]):
-    PIL = attempt_import_or_raise("PIL", "pillow")
+    PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
     if isinstance(input_data, str):
         if is_valid_url(input_data):
-            content = {"type": "image_url", "image_url": input_data}
+            if is_video_url(input_data):
+                content = {"type": "video_url", "video_url": input_data}
+            else:
+                content = {"type": "image_url", "image_url": input_data}
         else:
             content = {"type": "text", "text": input_data}
-    elif isinstance(input_data, PIL.Image.Image):
+    elif isinstance(input_data, PIL_Image.Image):
         buffered = BytesIO()
         input_data.save(buffered, format="JPEG")
         img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
@@ -61,7 +82,7 @@ def transform_input(input_data: Union[str, bytes, Path]):
             "image_base64": "data:image/jpeg;base64," + img_str,
         }
     elif isinstance(input_data, bytes):
-        img = PIL.Image.open(BytesIO(input_data))
+        img = PIL_Image.open(BytesIO(input_data))
         buffered = BytesIO()
         img.save(buffered, format="JPEG")
         img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
@@ -70,14 +91,24 @@ def transform_input(input_data: Union[str, bytes, Path]):
             "image_base64": "data:image/jpeg;base64," + img_str,
         }
     elif isinstance(input_data, Path):
-        img = PIL.Image.open(input_data)
-        buffered = BytesIO()
-        img.save(buffered, format="JPEG")
-        img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
-        content = {
-            "type": "image_base64",
-            "image_base64": "data:image/jpeg;base64," + img_str,
-        }
+        if is_video_path(input_data):
+            # Read video file and encode as base64
+            with open(input_data, "rb") as f:
+                video_bytes = f.read()
+            video_str = base64.b64encode(video_bytes).decode("utf-8")
+            content = {
+                "type": "video_base64",
+                "video_base64": video_str,
+            }
+        else:
+            img = PIL_Image.open(input_data)
+            buffered = BytesIO()
+            img.save(buffered, format="JPEG")
+            img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
+            content = {
+                "type": "image_base64",
+                "image_base64": "data:image/jpeg;base64," + img_str,
+            }
     else:
         raise ValueError("Each input should be either str, bytes, Path or Image.")
 
@@ -88,9 +119,11 @@ def sanitize_multimodal_input(inputs: Union[TEXT, IMAGES]) -> List[Any]:
     """
     Sanitize the input to the embedding function.
     """
-    PIL = attempt_import_or_raise("PIL", "pillow")
-    if isinstance(inputs, (str, bytes, Path, PIL.Image.Image)):
+    PIL_Image = attempt_import_or_raise("PIL.Image", "pillow")
+    if isinstance(inputs, (str, bytes, Path, PIL_Image.Image)):
         inputs = [inputs]
+    elif isinstance(inputs, list):
+        pass  # Already a list, use as-is
     elif isinstance(inputs, pa.Array):
         inputs = inputs.to_pylist()
     elif isinstance(inputs, pa.ChunkedArray):
@@ -100,7 +133,7 @@ def sanitize_multimodal_input(inputs: Union[TEXT, IMAGES]) -> List[Any]:
             f"Input type {type(inputs)} not allowed with multimodal model."
         )
 
-    if not all(isinstance(x, (str, bytes, Path, PIL.Image.Image)) for x in inputs):
+    if not all(isinstance(x, (str, bytes, Path, PIL_Image.Image)) for x in inputs):
         raise ValueError("Each input should be either str, bytes, Path or Image.")
 
     return [transform_input(i) for i in inputs]
@@ -137,17 +170,25 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
     name: str
         The name of the model to use. List of acceptable models:
 
+            * voyage-4 (1024 dims, general-purpose and multilingual retrieval)
+            * voyage-4-lite (1024 dims, optimized for latency and cost)
+            * voyage-4-large (1024 dims, best retrieval quality)
             * voyage-context-3
             * voyage-3.5
             * voyage-3.5-lite
             * voyage-3
             * voyage-3-lite
             * voyage-multimodal-3
+            * voyage-multimodal-3.5
             * voyage-finance-2
             * voyage-multilingual-2
             * voyage-law-2
             * voyage-code-2
 
+    output_dimension: int, optional
+        The output dimension for models that support flexible dimensions.
+        Currently only voyage-multimodal-3.5 supports this feature.
+        Valid options: 256, 512, 1024 (default), 2048.
 
     Examples
     --------
@@ -175,8 +216,14 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
     """
 
     name: str
+    output_dimension: Optional[int] = None
     client: ClassVar = None
+    _FLEXIBLE_DIM_MODELS: ClassVar[list] = ["voyage-multimodal-3.5"]
+    _VALID_DIMENSIONS: ClassVar[list] = [256, 512, 1024, 2048]
     text_embedding_models: list = [
+        "voyage-4",
+        "voyage-4-lite",
+        "voyage-4-large",
         "voyage-3.5",
         "voyage-3.5-lite",
         "voyage-3",
@@ -186,7 +233,7 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
         "voyage-law-2",
         "voyage-code-2",
     ]
-    multimodal_embedding_models: list = ["voyage-multimodal-3"]
+    multimodal_embedding_models: list = ["voyage-multimodal-3", "voyage-multimodal-3.5"]
     contextual_embedding_models: list = ["voyage-context-3"]
 
     def _is_multimodal_model(self, model_name: str):
@@ -198,11 +245,25 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
         return model_name in self.contextual_embedding_models or "context" in model_name
 
     def ndims(self):
+        # Handle flexible dimension models
+        if self.name in self._FLEXIBLE_DIM_MODELS:
+            if self.output_dimension is not None:
+                if self.output_dimension not in self._VALID_DIMENSIONS:
+                    raise ValueError(
+                        f"Invalid output_dimension {self.output_dimension} "
+                        f"for {self.name}. Valid options: {self._VALID_DIMENSIONS}"
+                    )
+                return self.output_dimension
+            return 1024  # default dimension
+
         if self.name == "voyage-3-lite":
             return 512
         elif self.name == "voyage-code-2":
             return 1536
         elif self.name in [
+            "voyage-4",
+            "voyage-4-lite",
+            "voyage-4-large",
             "voyage-context-3",
             "voyage-3.5",
             "voyage-3.5-lite",
@@ -211,12 +272,17 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
             "voyage-finance-2",
             "voyage-multilingual-2",
             "voyage-law-2",
-            "voyage-multimodal-3",
         ]:
             return 1024
         else:
             raise ValueError(f"Model {self.name} not supported")
 
+    def _get_multimodal_kwargs(self, **kwargs):
+        """Get kwargs for multimodal embed call, including output_dimension if set."""
+        if self.name in self._FLEXIBLE_DIM_MODELS and self.output_dimension is not None:
+            kwargs["output_dimension"] = self.output_dimension
+        return kwargs
+
     def compute_query_embeddings(
         self, query: Union[str, "PIL.Image.Image"], *args, **kwargs
     ) -> List[np.ndarray]:
@@ -234,6 +300,7 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
         """
         client = VoyageAIEmbeddingFunction._get_client()
         if self._is_multimodal_model(self.name):
+            kwargs = self._get_multimodal_kwargs(**kwargs)
             result = client.multimodal_embed(
                 inputs=[[query]], model=self.name, input_type="query", **kwargs
             )
@@ -275,6 +342,7 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
             )
             if has_images:
                 # Use non-batched API for images
+                kwargs = self._get_multimodal_kwargs(**kwargs)
                 result = client.multimodal_embed(
                     inputs=sanitized, model=self.name, input_type="document", **kwargs
                 )
@@ -357,6 +425,7 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
             callable: A function that takes a batch of texts and returns embeddings.
         """
         if self._is_multimodal_model(self.name):
+            multimodal_kwargs = self._get_multimodal_kwargs(**kwargs)
 
             def embed_batch(batch: List[str]) -> List[np.array]:
                 batch_inputs = sanitize_multimodal_input(batch)
@@ -364,7 +433,7 @@ class VoyageAIEmbeddingFunction(EmbeddingFunction):
                     inputs=batch_inputs,
                     model=self.name,
                     input_type=input_type,
-                    **kwargs,
+                    **multimodal_kwargs,
                 )
                 return result.embeddings
 

python/python/lancedb/pydantic.py

@@ -275,7 +275,7 @@ def _py_type_to_arrow_type(py_type: Type[Any], field: FieldInfo) -> pa.DataType:
         return pa.timestamp("us", tz=tz)
     elif getattr(py_type, "__origin__", None) in (list, tuple):
         child = py_type.__args__[0]
-        return pa.list_(_py_type_to_arrow_type(child, field))
+        return _pydantic_list_child_to_arrow(child, field)
     raise TypeError(
         f"Converting Pydantic type to Arrow Type: unsupported type {py_type}."
     )
@@ -298,12 +298,18 @@ else:
 
 
 def _pydantic_type_to_arrow_type(tp: Any, field: FieldInfo) -> pa.DataType:
+    def _safe_issubclass(candidate: Any, base: type) -> bool:
+        try:
+            return issubclass(candidate, base)
+        except TypeError:
+            return False
+
     if inspect.isclass(tp):
-        if issubclass(tp, pydantic.BaseModel):
+        if _safe_issubclass(tp, pydantic.BaseModel):
             # Struct
             fields = _pydantic_model_to_fields(tp)
             return pa.struct(fields)
-        if issubclass(tp, FixedSizeListMixin):
+        if _safe_issubclass(tp, FixedSizeListMixin):
             if getattr(tp, "is_multi_vector", lambda: False)():
                 return pa.list_(pa.list_(tp.value_arrow_type(), tp.dim()))
             # For regular Vector
@@ -311,45 +317,67 @@ def _pydantic_type_to_arrow_type(tp: Any, field: FieldInfo) -> pa.DataType:
     return _py_type_to_arrow_type(tp, field)
 
 
+def _pydantic_list_child_to_arrow(child: Any, field: FieldInfo) -> pa.DataType:
+    unwrapped = _unwrap_optional_annotation(child)
+    if unwrapped is not None:
+        return pa.list_(
+            pa.field("item", _pydantic_type_to_arrow_type(unwrapped, field), True)
+        )
+    return pa.list_(_pydantic_type_to_arrow_type(child, field))
+
+
+def _unwrap_optional_annotation(annotation: Any) -> Any | None:
+    if isinstance(annotation, (_GenericAlias, GenericAlias)):
+        origin = annotation.__origin__
+        args = annotation.__args__
+        if origin == Union:
+            non_none = [arg for arg in args if arg is not type(None)]
+            if len(non_none) == 1 and len(non_none) != len(args):
+                return non_none[0]
+    elif sys.version_info >= (3, 10) and isinstance(annotation, types.UnionType):
+        args = annotation.__args__
+        non_none = [arg for arg in args if arg is not type(None)]
+        if len(non_none) == 1 and len(non_none) != len(args):
+            return non_none[0]
+    return None
+
+
 def _pydantic_to_arrow_type(field: FieldInfo) -> pa.DataType:
     """Convert a Pydantic FieldInfo to Arrow DataType"""
+    unwrapped = _unwrap_optional_annotation(field.annotation)
+    if unwrapped is not None:
+        return _pydantic_type_to_arrow_type(unwrapped, field)
     if isinstance(field.annotation, (_GenericAlias, GenericAlias)):
         origin = field.annotation.__origin__
         args = field.annotation.__args__
 
         if origin is list:
             child = args[0]
-            return pa.list_(_py_type_to_arrow_type(child, field))
-        elif origin == Union:
-            if len(args) == 2 and args[1] is type(None):
-                return _pydantic_type_to_arrow_type(args[0], field)
-    elif sys.version_info >= (3, 10) and isinstance(field.annotation, types.UnionType):
-        args = field.annotation.__args__
-        if len(args) == 2:
-            for typ in args:
-                if typ is type(None):
-                    continue
-                return _py_type_to_arrow_type(typ, field)
+            return _pydantic_list_child_to_arrow(child, field)
     return _pydantic_type_to_arrow_type(field.annotation, field)
 
 
 def is_nullable(field: FieldInfo) -> bool:
     """Check if a Pydantic FieldInfo is nullable."""
+    if _unwrap_optional_annotation(field.annotation) is not None:
+        return True
     if isinstance(field.annotation, (_GenericAlias, GenericAlias)):
         origin = field.annotation.__origin__
         args = field.annotation.__args__
         if origin == Union:
-            if len(args) == 2 and args[1] is type(None):
+            if any(typ is type(None) for typ in args):
                 return True
     elif sys.version_info >= (3, 10) and isinstance(field.annotation, types.UnionType):
         args = field.annotation.__args__
         for typ in args:
             if typ is type(None):
                 return True
-    elif inspect.isclass(field.annotation) and issubclass(
-        field.annotation, FixedSizeListMixin
-    ):
-        return field.annotation.nullable()
+    elif inspect.isclass(field.annotation):
+        try:
+            if issubclass(field.annotation, FixedSizeListMixin):
+                return field.annotation.nullable()
+        except TypeError:
+            return False
     return False
 
 

python/python/lancedb/remote/db.py

@@ -384,6 +384,7 @@ class RemoteDBConnection(DBConnection):
         on_bad_vectors: str = "error",
         fill_value: float = 0.0,
         mode: Optional[str] = None,
+        exist_ok: bool = False,
         embedding_functions: Optional[List[EmbeddingFunctionConfig]] = None,
         *,
         namespace: Optional[List[str]] = None,
@@ -412,6 +413,12 @@ class RemoteDBConnection(DBConnection):
             - pyarrow.Schema
 
             - [LanceModel][lancedb.pydantic.LanceModel]
+        mode: str, default "create"
+            The mode to use when creating the table.
+            Can be either "create", "overwrite", or "exist_ok".
+        exist_ok: bool, default False
+            If exist_ok is True, and mode is None or "create", mode will be changed
+            to "exist_ok".
         on_bad_vectors: str, default "error"
             What to do if any of the vectors are not the same size or contains NaNs.
             One of "error", "drop", "fill".
@@ -483,6 +490,11 @@ class RemoteDBConnection(DBConnection):
         LanceTable(table4)
 
         """
+        if exist_ok:
+            if mode == "create":
+                mode = "exist_ok"
+            elif not mode:
+                mode = "exist_ok"
         if namespace is None:
             namespace = []
         validate_table_name(name)

python/python/lancedb/remote/table.py

@@ -18,7 +18,17 @@ from lancedb._lancedb import (
     UpdateResult,
 )
 from lancedb.embeddings.base import EmbeddingFunctionConfig
-from lancedb.index import FTS, BTree, Bitmap, HnswSq, IvfFlat, IvfPq, IvfSq, LabelList
+from lancedb.index import (
+    FTS,
+    BTree,
+    Bitmap,
+    HnswSq,
+    IvfFlat,
+    IvfPq,
+    IvfRq,
+    IvfSq,
+    LabelList,
+)
 from lancedb.remote.db import LOOP
 import pyarrow as pa
 
@@ -265,6 +275,12 @@ class RemoteTable(Table):
                 num_sub_vectors=num_sub_vectors,
                 num_bits=num_bits,
             )
+        elif index_type == "IVF_RQ":
+            config = IvfRq(
+                distance_type=metric,
+                num_partitions=num_partitions,
+                num_bits=num_bits,
+            )
         elif index_type == "IVF_SQ":
             config = IvfSq(distance_type=metric, num_partitions=num_partitions)
         elif index_type == "IVF_HNSW_PQ":
@@ -279,7 +295,8 @@ class RemoteTable(Table):
         else:
             raise ValueError(
                 f"Unknown vector index type: {index_type}. Valid options are"
-                " 'IVF_FLAT', 'IVF_SQ', 'IVF_PQ', 'IVF_HNSW_PQ', 'IVF_HNSW_SQ'"
+                " 'IVF_FLAT', 'IVF_PQ', 'IVF_RQ', 'IVF_SQ',"
+                " 'IVF_HNSW_PQ', 'IVF_HNSW_SQ'"
             )
 
         LOOP.run(
@@ -638,6 +655,14 @@ class RemoteTable(Table):
     def stats(self):
         return LOOP.run(self._table.stats())
 
+    @property
+    def uri(self) -> str:
+        """The table URI (storage location).
+
+        For remote tables, this fetches the location from the server via describe.
+        """
+        return LOOP.run(self._table.uri())
+
     def take_offsets(self, offsets: list[int]) -> LanceTakeQueryBuilder:
         return LanceTakeQueryBuilder(self._table.take_offsets(offsets))
 

python/python/lancedb/table.py

@@ -684,6 +684,24 @@ class Table(ABC):
         """
         raise NotImplementedError
 
+    def to_lance(self, **kwargs) -> lance.LanceDataset:
+        """Return the table as a lance.LanceDataset.
+
+        Returns
+        -------
+        lance.LanceDataset
+        """
+        raise NotImplementedError
+
+    def to_polars(self, **kwargs) -> "pl.DataFrame":
+        """Return the table as a polars.DataFrame.
+
+        Returns
+        -------
+        polars.DataFrame
+        """
+        raise NotImplementedError
+
     def create_index(
         self,
         metric="l2",
@@ -2200,6 +2218,10 @@ class LanceTable(Table):
     def stats(self) -> TableStatistics:
         return LOOP.run(self._table.stats())
 
+    @property
+    def uri(self) -> str:
+        return LOOP.run(self._table.uri())
+
     def create_scalar_index(
         self,
         column: str,
@@ -3208,7 +3230,27 @@ def _infer_target_schema(
             if pa.types.is_floating(field.type.value_type):
                 target_type = pa.list_(pa.float32(), dim)
             elif pa.types.is_integer(field.type.value_type):
-                target_type = pa.list_(pa.uint8(), dim)
+                values = peeked.column(i)
+
+                if isinstance(values, pa.ChunkedArray):
+                    values = values.combine_chunks()
+
+                flattened = values.flatten()
+                valid_count = pc.count(flattened, mode="only_valid").as_py()
+
+                if valid_count == 0:
+                    target_type = pa.list_(pa.uint8(), dim)
+                else:
+                    min_max = pc.min_max(flattened)
+                    min_value = min_max["min"].as_py()
+                    max_value = min_max["max"].as_py()
+
+                    if (min_value is not None and min_value < 0) or (
+                        max_value is not None and max_value > 255
+                    ):
+                        target_type = pa.list_(pa.float32(), dim)
+                    else:
+                        target_type = pa.list_(pa.uint8(), dim)
             else:
                 continue  # Skip non-numeric types
 
@@ -3568,6 +3610,20 @@ class AsyncTable:
         """
         return await self._inner.stats()
 
+    async def uri(self) -> str:
+        """
+        Get the table URI (storage location).
+
+        For remote tables, this fetches the location from the server via describe.
+        For local tables, this returns the dataset URI.
+
+        Returns
+        -------
+        str
+            The full storage location of the table (e.g., S3/GCS path).
+        """
+        return await self._inner.uri()
+
     async def add(
         self,
         data: DATA,

python/python/tests/conftest.py

@@ -2,12 +2,27 @@
 # SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
 from datetime import timedelta
+
 from lancedb.db import AsyncConnection, DBConnection
 import lancedb
 import pytest
 import pytest_asyncio
 
 
+def pandas_string_type():
+    """Return the PyArrow string type that pandas uses for string columns.
+
+    pandas 3.0+ uses large_string for string columns, pandas 2.x uses string.
+    """
+    import pandas as pd
+    import pyarrow as pa
+
+    version = tuple(int(x) for x in pd.__version__.split(".")[:2])
+    if version >= (3, 0):
+        return pa.large_utf8()
+    return pa.utf8()
+
+
 # Use an in-memory database for most tests.
 @pytest.fixture
 def mem_db() -> DBConnection:

python/python/tests/test_db.py

@@ -268,6 +268,8 @@ async def test_create_table_from_iterator_async(mem_db_async: lancedb.AsyncConne
 
 
 def test_create_exist_ok(tmp_db: lancedb.DBConnection):
+    from conftest import pandas_string_type
+
     data = pd.DataFrame(
         {
             "vector": [[3.1, 4.1], [5.9, 26.5]],
@@ -286,10 +288,11 @@ def test_create_exist_ok(tmp_db: lancedb.DBConnection):
     assert tbl.schema == tbl2.schema
     assert len(tbl) == len(tbl2)
 
+    # pandas 3.0+ uses large_string, pandas 2.x uses string
     schema = pa.schema(
         [
             pa.field("vector", pa.list_(pa.float32(), list_size=2)),
-            pa.field("item", pa.utf8()),
+            pa.field("item", pandas_string_type()),
             pa.field("price", pa.float64()),
         ]
     )
@@ -299,7 +302,7 @@ def test_create_exist_ok(tmp_db: lancedb.DBConnection):
     bad_schema = pa.schema(
         [
             pa.field("vector", pa.list_(pa.float32(), list_size=2)),
-            pa.field("item", pa.utf8()),
+            pa.field("item", pandas_string_type()),
             pa.field("price", pa.float64()),
             pa.field("extra", pa.float32()),
         ]
@@ -365,6 +368,8 @@ async def test_create_mode_async(tmp_db_async: lancedb.AsyncConnection):
 
 @pytest.mark.asyncio
 async def test_create_exist_ok_async(tmp_db_async: lancedb.AsyncConnection):
+    from conftest import pandas_string_type
+
     data = pd.DataFrame(
         {
             "vector": [[3.1, 4.1], [5.9, 26.5]],
@@ -382,10 +387,11 @@ async def test_create_exist_ok_async(tmp_db_async: lancedb.AsyncConnection):
     assert tbl.name == tbl2.name
     assert await tbl.schema() == await tbl2.schema()
 
+    # pandas 3.0+ uses large_string, pandas 2.x uses string
     schema = pa.schema(
         [
             pa.field("vector", pa.list_(pa.float32(), list_size=2)),
-            pa.field("item", pa.utf8()),
+            pa.field("item", pandas_string_type()),
             pa.field("price", pa.float64()),
         ]
     )
@@ -595,6 +601,8 @@ def test_open_table_sync(tmp_db: lancedb.DBConnection):
 
 @pytest.mark.asyncio
 async def test_open_table(tmp_path):
+    from conftest import pandas_string_type
+
     db = await lancedb.connect_async(tmp_path)
     data = pd.DataFrame(
         {
@@ -614,10 +622,11 @@ async def test_open_table(tmp_path):
         )
         is not None
     )
+    # pandas 3.0+ uses large_string, pandas 2.x uses string
     assert await tbl.schema() == pa.schema(
         {
             "vector": pa.list_(pa.float32(), list_size=2),
-            "item": pa.utf8(),
+            "item": pandas_string_type(),
             "price": pa.float64(),
         }
     )

python/python/tests/test_embeddings_slow.py

@@ -517,19 +517,36 @@ def test_ollama_embedding(tmp_path):
 @pytest.mark.skipif(
     os.environ.get("VOYAGE_API_KEY") is None, reason="VOYAGE_API_KEY not set"
 )
-def test_voyageai_embedding_function():
-    voyageai = get_registry().get("voyageai").create(name="voyage-3", max_retries=0)
+@pytest.mark.parametrize(
+    "model_name,expected_dims",
+    [
+        ("voyage-3", 1024),
+        ("voyage-4", 1024),
+        ("voyage-4-lite", 1024),
+        ("voyage-4-large", 1024),
+    ],
+)
+def test_voyageai_embedding_function(model_name, expected_dims, tmp_path):
+    """Integration test for VoyageAI text embedding models with real API calls."""
+    voyageai = get_registry().get("voyageai").create(name=model_name, max_retries=0)
 
     class TextModel(LanceModel):
         text: str = voyageai.SourceField()
         vector: Vector(voyageai.ndims()) = voyageai.VectorField()
 
     df = pd.DataFrame({"text": ["hello world", "goodbye world"]})
-    db = lancedb.connect("~/lancedb")
+    db = lancedb.connect(tmp_path)
     tbl = db.create_table("test", schema=TextModel, mode="overwrite")
 
     tbl.add(df)
     assert len(tbl.to_pandas()["vector"][0]) == voyageai.ndims()
+    assert voyageai.ndims() == expected_dims, (
+        f"{model_name} should have {expected_dims} dimensions"
+    )
+
+    # Test search functionality
+    result = tbl.search("hello").limit(1).to_pandas()
+    assert result["text"][0] == "hello world"
 
 
 @pytest.mark.slow
@@ -613,6 +630,133 @@ def test_voyageai_multimodal_embedding_text_function():
     assert len(tbl.to_pandas()["vector"][0]) == voyageai.ndims()
 
 
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("VOYAGE_API_KEY") is None, reason="VOYAGE_API_KEY not set"
+)
+def test_voyageai_multimodal_35_embedding_function():
+    """Test voyage-multimodal-3.5 model with text input."""
+    voyageai = (
+        get_registry()
+        .get("voyageai")
+        .create(name="voyage-multimodal-3.5", max_retries=0)
+    )
+
+    class TextModel(LanceModel):
+        text: str = voyageai.SourceField()
+        vector: Vector(voyageai.ndims()) = voyageai.VectorField()
+
+    df = pd.DataFrame({"text": ["hello world", "goodbye world"]})
+    db = lancedb.connect("~/lancedb")
+    tbl = db.create_table("test_multimodal_35", schema=TextModel, mode="overwrite")
+
+    tbl.add(df)
+    assert len(tbl.to_pandas()["vector"][0]) == voyageai.ndims()
+    assert voyageai.ndims() == 1024
+
+
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("VOYAGE_API_KEY") is None, reason="VOYAGE_API_KEY not set"
+)
+def test_voyageai_multimodal_35_flexible_dimensions():
+    """Test voyage-multimodal-3.5 model with custom output dimension."""
+    voyageai = (
+        get_registry()
+        .get("voyageai")
+        .create(name="voyage-multimodal-3.5", output_dimension=512, max_retries=0)
+    )
+
+    class TextModel(LanceModel):
+        text: str = voyageai.SourceField()
+        vector: Vector(voyageai.ndims()) = voyageai.VectorField()
+
+    assert voyageai.ndims() == 512
+
+    df = pd.DataFrame({"text": ["hello world", "goodbye world"]})
+    db = lancedb.connect("~/lancedb")
+    tbl = db.create_table("test_multimodal_35_dim", schema=TextModel, mode="overwrite")
+
+    tbl.add(df)
+    assert len(tbl.to_pandas()["vector"][0]) == 512
+
+
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("VOYAGE_API_KEY") is None, reason="VOYAGE_API_KEY not set"
+)
+def test_voyageai_multimodal_35_image_embedding():
+    """Test voyage-multimodal-3.5 model with image input."""
+    voyageai = (
+        get_registry()
+        .get("voyageai")
+        .create(name="voyage-multimodal-3.5", max_retries=0)
+    )
+
+    class Images(LanceModel):
+        label: str
+        image_uri: str = voyageai.SourceField()
+        vector: Vector(voyageai.ndims()) = voyageai.VectorField()
+
+    db = lancedb.connect("~/lancedb")
+    table = db.create_table(
+        "test_multimodal_35_images", schema=Images, mode="overwrite"
+    )
+    labels = ["cat", "dog"]
+    uris = [
+        "http://farm1.staticflickr.com/53/167798175_7c7845bbbd_z.jpg",
+        "http://farm9.staticflickr.com/8387/8602747737_2e5c2a45d4_z.jpg",
+    ]
+    table.add(pd.DataFrame({"label": labels, "image_uri": uris}))
+    assert len(table.to_pandas()["vector"][0]) == voyageai.ndims()
+    assert voyageai.ndims() == 1024
+
+
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("VOYAGE_API_KEY") is None, reason="VOYAGE_API_KEY not set"
+)
+@pytest.mark.parametrize("dimension", [256, 512, 1024, 2048])
+def test_voyageai_multimodal_35_all_dimensions(dimension):
+    """Test voyage-multimodal-3.5 model with all valid output dimensions."""
+    voyageai = (
+        get_registry()
+        .get("voyageai")
+        .create(name="voyage-multimodal-3.5", output_dimension=dimension, max_retries=0)
+    )
+
+    assert voyageai.ndims() == dimension
+
+    class TextModel(LanceModel):
+        text: str = voyageai.SourceField()
+        vector: Vector(voyageai.ndims()) = voyageai.VectorField()
+
+    df = pd.DataFrame({"text": ["hello world"]})
+    db = lancedb.connect("~/lancedb")
+    tbl = db.create_table(
+        f"test_multimodal_35_dim_{dimension}", schema=TextModel, mode="overwrite"
+    )
+
+    tbl.add(df)
+    assert len(tbl.to_pandas()["vector"][0]) == dimension
+
+
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("VOYAGE_API_KEY") is None, reason="VOYAGE_API_KEY not set"
+)
+def test_voyageai_multimodal_35_invalid_dimension():
+    """Test voyage-multimodal-3.5 model raises error for invalid output dimension."""
+    with pytest.raises(ValueError, match="Invalid output_dimension"):
+        voyageai = (
+            get_registry()
+            .get("voyageai")
+            .create(name="voyage-multimodal-3.5", output_dimension=999, max_retries=0)
+        )
+        # ndims() is where the validation happens
+        voyageai.ndims()
+
+
 @pytest.mark.slow
 @pytest.mark.skipif(
     importlib.util.find_spec("colpali_engine") is None,

python/python/tests/test_namespace_integration.py

@@ -26,6 +26,8 @@ import pytest
 from lance_namespace import (
     CreateEmptyTableRequest,
     CreateEmptyTableResponse,
+    DeclareTableRequest,
+    DeclareTableResponse,
     DescribeTableRequest,
     DescribeTableResponse,
     LanceNamespace,
@@ -160,6 +162,19 @@ class TrackingNamespace(LanceNamespace):
 
         return modified
 
+    def declare_table(self, request: DeclareTableRequest) -> DeclareTableResponse:
+        """Track declare_table calls and inject rotating credentials."""
+        with self.lock:
+            self.create_call_count += 1
+            count = self.create_call_count
+
+        response = self.inner.declare_table(request)
+        response.storage_options = self._modify_storage_options(
+            response.storage_options, count
+        )
+
+        return response
+
     def create_empty_table(
         self, request: CreateEmptyTableRequest
     ) -> CreateEmptyTableResponse:

python/python/tests/test_permutation.py

@@ -438,11 +438,15 @@ def test_filter_with_splits(mem_db):
     row_count = permutation_tbl.count_rows()
     assert row_count == 67
 
-    data = permutation_tbl.search(None).to_arrow().to_pydict()
+    # Verify the permutation table only contains row_id and split_id
+    assert set(permutation_tbl.schema.names) == {"row_id", "split_id"}
+
+    row_ids = permutation_tbl.search(None).to_arrow().to_pydict()["row_id"]
+    data = tbl.take_row_ids(row_ids).to_arrow().to_pydict()
     categories = data["category"]
 
     # All categories should be A or B
-    assert all(cat in ["A", "B"] for cat in categories)
+    assert all(cat in ("A", "B") for cat in categories)
 
 
 def test_filter_with_shuffle(mem_db):

python/python/tests/test_pydantic.py

@@ -2,7 +2,6 @@
 # SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
 import json
-import sys
 from datetime import date, datetime
 from typing import List, Optional, Tuple
 
@@ -20,10 +19,6 @@ from pydantic import BaseModel
 from pydantic import Field
 
 
-@pytest.mark.skipif(
-    sys.version_info < (3, 9),
-    reason="using native type alias requires python3.9 or higher",
-)
 def test_pydantic_to_arrow():
     class StructModel(pydantic.BaseModel):
         a: str
@@ -83,10 +78,6 @@ def test_pydantic_to_arrow():
     assert schema == expect_schema
 
 
-@pytest.mark.skipif(
-    sys.version_info < (3, 10),
-    reason="using | type syntax requires python3.10 or higher",
-)
 def test_optional_types_py310():
     class TestModel(pydantic.BaseModel):
         a: str | None
@@ -105,10 +96,233 @@ def test_optional_types_py310():
     assert schema == expect_schema
 
 
-@pytest.mark.skipif(
-    sys.version_info > (3, 8),
-    reason="using native type alias requires python3.9 or higher",
-)
+def test_optional_structs():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        split: SplitInfo | None = None
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "split",
+                pa.struct(
+                    [
+                        pa.field("start_frame", pa.int64(), False),
+                        pa.field("end_frame", pa.int64(), False),
+                    ]
+                ),
+                True,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
+def test_optional_struct_list_py310():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        splits: list[SplitInfo] | None = None
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "splits",
+                pa.list_(
+                    pa.struct(
+                        [
+                            pa.field("start_frame", pa.int64(), False),
+                            pa.field("end_frame", pa.int64(), False),
+                        ]
+                    )
+                ),
+                True,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
+def test_nested_struct_list():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        splits: list[SplitInfo]
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "splits",
+                pa.list_(
+                    pa.struct(
+                        [
+                            pa.field("start_frame", pa.int64(), False),
+                            pa.field("end_frame", pa.int64(), False),
+                        ]
+                    )
+                ),
+                False,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
+def test_nested_struct_list_optional():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        splits: Optional[list[SplitInfo]] = None
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "splits",
+                pa.list_(
+                    pa.struct(
+                        [
+                            pa.field("start_frame", pa.int64(), False),
+                            pa.field("end_frame", pa.int64(), False),
+                        ]
+                    )
+                ),
+                True,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
+def test_nested_struct_list_optional_items():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        splits: list[Optional[SplitInfo]]
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "splits",
+                pa.list_(
+                    pa.field(
+                        "item",
+                        pa.struct(
+                            [
+                                pa.field("start_frame", pa.int64(), False),
+                                pa.field("end_frame", pa.int64(), False),
+                            ]
+                        ),
+                        True,
+                    )
+                ),
+                False,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
+def test_nested_struct_list_optional_container_and_items():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        splits: Optional[list[Optional[SplitInfo]]] = None
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "splits",
+                pa.list_(
+                    pa.field(
+                        "item",
+                        pa.struct(
+                            [
+                                pa.field("start_frame", pa.int64(), False),
+                                pa.field("end_frame", pa.int64(), False),
+                            ]
+                        ),
+                        True,
+                    )
+                ),
+                True,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
+def test_nested_struct_list_optional_items_pep604():
+    class SplitInfo(pydantic.BaseModel):
+        start_frame: int
+        end_frame: int
+
+    class TestModel(pydantic.BaseModel):
+        id: str
+        splits: list[SplitInfo | None]
+
+    schema = pydantic_to_schema(TestModel)
+
+    expect_schema = pa.schema(
+        [
+            pa.field("id", pa.utf8(), False),
+            pa.field(
+                "splits",
+                pa.list_(
+                    pa.field(
+                        "item",
+                        pa.struct(
+                            [
+                                pa.field("start_frame", pa.int64(), False),
+                                pa.field("end_frame", pa.int64(), False),
+                            ]
+                        ),
+                        True,
+                    )
+                ),
+                False,
+            ),
+        ]
+    )
+    assert schema == expect_schema
+
+
 def test_pydantic_to_arrow_py38():
     class StructModel(pydantic.BaseModel):
         a: str

python/python/tests/test_remote_db.py

@@ -8,7 +8,7 @@ import http.server
 import json
 import threading
 import time
-from unittest.mock import MagicMock
+from unittest.mock import MagicMock, patch
 import uuid
 from packaging.version import Version
 
@@ -168,6 +168,42 @@ def test_table_len_sync():
         assert len(table) == 1
 
 
+def test_create_table_exist_ok():
+    def handler(request):
+        if request.path == "/v1/table/test/create/?mode=exist_ok":
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            request.wfile.write(b"{}")
+        else:
+            request.send_response(404)
+            request.end_headers()
+
+    with mock_lancedb_connection(handler) as db:
+        table = db.create_table("test", [{"id": 1}], exist_ok=True)
+        assert table is not None
+
+    with mock_lancedb_connection(handler) as db:
+        table = db.create_table("test", [{"id": 1}], mode="create", exist_ok=True)
+        assert table is not None
+
+
+def test_create_table_exist_ok_with_mode_overwrite():
+    def handler(request):
+        if request.path == "/v1/table/test/create/?mode=overwrite":
+            request.send_response(200)
+            request.send_header("Content-Type", "application/json")
+            request.end_headers()
+            request.wfile.write(b"{}")
+        else:
+            request.send_response(404)
+            request.end_headers()
+
+    with mock_lancedb_connection(handler) as db:
+        table = db.create_table("test", [{"id": 1}], mode="overwrite", exist_ok=True)
+        assert table is not None
+
+
 @pytest.mark.asyncio
 async def test_http_error():
     request_id_holder = {"request_id": None}
@@ -1167,3 +1203,22 @@ async def test_header_provider_overrides_static_headers():
         extra_headers={"X-API-Key": "static-key", "X-Extra": "extra-value"},
     ) as db:
         await db.table_names()
+
+
+@pytest.mark.parametrize("exception", [KeyboardInterrupt, SystemExit, GeneratorExit])
+def test_background_loop_cancellation(exception):
+    """Test that BackgroundEventLoop.run() cancels the future on interrupt."""
+    from lancedb.background_loop import BackgroundEventLoop
+
+    mock_future = MagicMock()
+    mock_future.result.side_effect = exception()
+
+    with (
+        patch.object(BackgroundEventLoop, "__init__", return_value=None),
+        patch("asyncio.run_coroutine_threadsafe", return_value=mock_future),
+    ):
+        loop = BackgroundEventLoop()
+        loop.loop = MagicMock()
+        with pytest.raises(exception):
+            loop.run(None)
+        mock_future.cancel.assert_called_once()

python/python/tests/test_s3_bucket_dots.py

@@ -0,0 +1,68 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+"""
+Tests for S3 bucket names containing dots.
+
+Related issue: https://github.com/lancedb/lancedb/issues/1898
+
+These tests validate the early error checking for S3 bucket names with dots.
+No actual S3 connection is made - validation happens before connection.
+"""
+
+import pytest
+import lancedb
+
+# Test URIs
+BUCKET_WITH_DOTS = "s3://my.bucket.name/path"
+BUCKET_WITH_DOTS_AND_REGION = ("s3://my.bucket.name", {"region": "us-east-1"})
+BUCKET_WITH_DOTS_AND_AWS_REGION = ("s3://my.bucket.name", {"aws_region": "us-east-1"})
+BUCKET_WITHOUT_DOTS = "s3://my-bucket/path"
+
+
+class TestS3BucketWithDotsSync:
+    """Tests for connect()."""
+
+    def test_bucket_with_dots_requires_region(self):
+        with pytest.raises(ValueError, match="contains dots"):
+            lancedb.connect(BUCKET_WITH_DOTS)
+
+    def test_bucket_with_dots_and_region_passes(self):
+        uri, opts = BUCKET_WITH_DOTS_AND_REGION
+        db = lancedb.connect(uri, storage_options=opts)
+        assert db is not None
+
+    def test_bucket_with_dots_and_aws_region_passes(self):
+        uri, opts = BUCKET_WITH_DOTS_AND_AWS_REGION
+        db = lancedb.connect(uri, storage_options=opts)
+        assert db is not None
+
+    def test_bucket_without_dots_passes(self):
+        db = lancedb.connect(BUCKET_WITHOUT_DOTS)
+        assert db is not None
+
+
+class TestS3BucketWithDotsAsync:
+    """Tests for connect_async()."""
+
+    @pytest.mark.asyncio
+    async def test_bucket_with_dots_requires_region(self):
+        with pytest.raises(ValueError, match="contains dots"):
+            await lancedb.connect_async(BUCKET_WITH_DOTS)
+
+    @pytest.mark.asyncio
+    async def test_bucket_with_dots_and_region_passes(self):
+        uri, opts = BUCKET_WITH_DOTS_AND_REGION
+        db = await lancedb.connect_async(uri, storage_options=opts)
+        assert db is not None
+
+    @pytest.mark.asyncio
+    async def test_bucket_with_dots_and_aws_region_passes(self):
+        uri, opts = BUCKET_WITH_DOTS_AND_AWS_REGION
+        db = await lancedb.connect_async(uri, storage_options=opts)
+        assert db is not None
+
+    @pytest.mark.asyncio
+    async def test_bucket_without_dots_passes(self):
+        db = await lancedb.connect_async(BUCKET_WITHOUT_DOTS)
+        assert db is not None

python/python/tests/test_table.py

@@ -46,6 +46,39 @@ def test_basic(mem_db: DBConnection):
     assert table.to_arrow() == expected_data
 
 
+def test_create_table_infers_large_int_vectors(mem_db: DBConnection):
+    data = [{"vector": [0, 300]}]
+
+    table = mem_db.create_table(
+        "int_vector_overflow", data=data, mode="overwrite", exist_ok=True
+    )
+
+    vector_field = table.schema.field("vector")
+    assert vector_field.type == pa.list_(pa.float32(), 2)
+
+    vector_column = table.to_arrow().column("vector")
+    assert vector_column.type == pa.list_(pa.float32(), 2)
+    assert vector_column.to_pylist() == [[0.0, 300.0]]
+
+
+@pytest.mark.asyncio
+async def test_create_table_async_infers_large_int_vectors(
+    mem_db_async: AsyncConnection,
+):
+    data = [{"vector": [256, 257]}]
+
+    table = await mem_db_async.create_table(
+        "int_vector_overflow_async", data=data, mode="overwrite", exist_ok=True
+    )
+
+    schema = await table.schema()
+    assert schema.field("vector").type == pa.list_(pa.float32(), 2)
+
+    vector_column = (await table.to_arrow()).column("vector")
+    assert vector_column.type == pa.list_(pa.float32(), 2)
+    assert vector_column.to_pylist() == [[256.0, 257.0]]
+
+
 def test_input_data_type(mem_db: DBConnection, tmp_path):
     schema = pa.schema(
         {
@@ -1934,3 +1967,9 @@ def test_add_table_with_empty_embeddings(tmp_path):
         on_bad_vectors="drop",
     )
     assert table.count_rows() == 1
+
+
+def test_table_uri(tmp_path):
+    db = lancedb.connect(tmp_path)
+    table = db.create_table("my_table", data=[{"x": 0}])
+    assert table.uri == str(tmp_path / "my_table.lance")

python/python/tests/test_util.py

@@ -528,12 +528,19 @@ def test_sanitize_data(
         else:
             expected_schema = schema
     else:
+        from conftest import pandas_string_type
+
+        # polars uses large_string, pandas 3.0+ uses large_string, others use string
+        if isinstance(data, pl.DataFrame):
+            text_type = pa.large_utf8()
+        elif isinstance(data, pd.DataFrame):
+            text_type = pandas_string_type()
+        else:
+            text_type = pa.string()
         expected_schema = pa.schema(
             {
                 "id": pa.int64(),
-                "text": pa.large_utf8()
-                if isinstance(data, pl.DataFrame)
-                else pa.string(),
+                "text": text_type,
                 "vector": pa.list_(pa.float32(), 10),
             }
         )

python/python/tests/test_voyageai_embeddings.py

@@ -0,0 +1,108 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+"""Unit tests for VoyageAI embedding function.
+
+These tests verify model registration and configuration without requiring API calls.
+"""
+
+import pytest
+from unittest.mock import MagicMock, patch
+
+from lancedb.embeddings import get_registry
+
+
+@pytest.fixture(autouse=True)
+def reset_voyageai_client():
+    """Reset VoyageAI client before and after each test to avoid state pollution."""
+    from lancedb.embeddings.voyageai import VoyageAIEmbeddingFunction
+
+    VoyageAIEmbeddingFunction.client = None
+    yield
+    VoyageAIEmbeddingFunction.client = None
+
+
+class TestVoyageAIModelRegistration:
+    """Tests for VoyageAI model registration and configuration."""
+
+    @pytest.fixture
+    def mock_voyageai_client(self):
+        """Mock VoyageAI client to avoid API calls."""
+        with patch.dict("os.environ", {"VOYAGE_API_KEY": "test-key"}):
+            with patch("lancedb.embeddings.voyageai.attempt_import_or_raise") as mock:
+                mock_client = MagicMock()
+                mock_voyageai = MagicMock()
+                mock_voyageai.Client.return_value = mock_client
+                mock.return_value = mock_voyageai
+                yield mock_client
+
+    def test_voyageai_registered(self):
+        """Test that VoyageAI is registered in the embedding function registry."""
+        registry = get_registry()
+        assert registry.get("voyageai") is not None
+
+    @pytest.mark.parametrize(
+        "model_name,expected_dims",
+        [
+            # Voyage-4 series (all 1024 dims)
+            ("voyage-4", 1024),
+            ("voyage-4-lite", 1024),
+            ("voyage-4-large", 1024),
+            # Voyage-3 series
+            ("voyage-3", 1024),
+            ("voyage-3-lite", 512),
+            # Domain-specific models
+            ("voyage-finance-2", 1024),
+            ("voyage-multilingual-2", 1024),
+            ("voyage-law-2", 1024),
+            ("voyage-code-2", 1536),
+            # Multimodal
+            ("voyage-multimodal-3", 1024),
+        ],
+    )
+    def test_model_dimensions(self, model_name, expected_dims, mock_voyageai_client):
+        """Test that each model returns the correct dimensions."""
+        registry = get_registry()
+        func = registry.get("voyageai").create(name=model_name)
+        assert func.ndims() == expected_dims, (
+            f"Model {model_name} should have {expected_dims} dimensions"
+        )
+
+    def test_unsupported_model_raises_error(self, mock_voyageai_client):
+        """Test that unsupported models raise ValueError."""
+        registry = get_registry()
+        func = registry.get("voyageai").create(name="unsupported-model")
+        with pytest.raises(ValueError, match="not supported"):
+            func.ndims()
+
+    @pytest.mark.parametrize(
+        "model_name",
+        [
+            "voyage-4",
+            "voyage-4-lite",
+            "voyage-4-large",
+        ],
+    )
+    def test_voyage4_models_are_text_models(self, model_name, mock_voyageai_client):
+        """Test that voyage-4 models are classified as text models (not multimodal)."""
+        registry = get_registry()
+        func = registry.get("voyageai").create(name=model_name)
+        assert not func._is_multimodal_model(model_name), (
+            f"{model_name} should be a text model, not multimodal"
+        )
+
+    def test_voyage4_models_in_text_embedding_list(self, mock_voyageai_client):
+        """Test that voyage-4 models are in the text_embedding_models list."""
+        registry = get_registry()
+        func = registry.get("voyageai").create(name="voyage-4")
+        assert "voyage-4" in func.text_embedding_models
+        assert "voyage-4-lite" in func.text_embedding_models
+        assert "voyage-4-large" in func.text_embedding_models
+
+    def test_voyage4_models_not_in_multimodal_list(self, mock_voyageai_client):
+        """Test that voyage-4 models are NOT in the multimodal_embedding_models list."""
+        registry = get_registry()
+        func = registry.get("voyageai").create(name="voyage-4")
+        assert "voyage-4" not in func.multimodal_embedding_models
+        assert "voyage-4-lite" not in func.multimodal_embedding_models
+        assert "voyage-4-large" not in func.multimodal_embedding_models

python/src/connection.rs

@@ -304,6 +304,7 @@ impl Connection {
                 },
                 page_token,
                 limit: limit.map(|l| l as i32),
+                ..Default::default()
             };
             let response = inner.list_namespaces(request).await.infer_error()?;
             Python::with_gil(|py| -> PyResult<Py<PyDict>> {
@@ -325,11 +326,12 @@ impl Connection {
         let inner = self_.get_inner()?.clone();
         let py = self_.py();
         future_into_py(py, async move {
-            use lance_namespace::models::{create_namespace_request, CreateNamespaceRequest};
-            let mode_enum = mode.and_then(|m| match m.to_lowercase().as_str() {
-                "create" => Some(create_namespace_request::Mode::Create),
-                "exist_ok" => Some(create_namespace_request::Mode::ExistOk),
-                "overwrite" => Some(create_namespace_request::Mode::Overwrite),
+            use lance_namespace::models::CreateNamespaceRequest;
+            // Mode is now a string field
+            let mode_str = mode.and_then(|m| match m.to_lowercase().as_str() {
+                "create" => Some("Create".to_string()),
+                "exist_ok" => Some("ExistOk".to_string()),
+                "overwrite" => Some("Overwrite".to_string()),
                 _ => None,
             });
             let request = CreateNamespaceRequest {
@@ -338,8 +340,9 @@ impl Connection {
                 } else {
                     Some(namespace)
                 },
-                mode: mode_enum,
+                mode: mode_str,
                 properties,
+                ..Default::default()
             };
             let response = inner.create_namespace(request).await.infer_error()?;
             Python::with_gil(|py| -> PyResult<Py<PyDict>> {
@@ -360,15 +363,16 @@ impl Connection {
         let inner = self_.get_inner()?.clone();
         let py = self_.py();
         future_into_py(py, async move {
-            use lance_namespace::models::{drop_namespace_request, DropNamespaceRequest};
-            let mode_enum = mode.and_then(|m| match m.to_uppercase().as_str() {
-                "SKIP" => Some(drop_namespace_request::Mode::Skip),
-                "FAIL" => Some(drop_namespace_request::Mode::Fail),
+            use lance_namespace::models::DropNamespaceRequest;
+            // Mode and Behavior are now string fields
+            let mode_str = mode.and_then(|m| match m.to_uppercase().as_str() {
+                "SKIP" => Some("Skip".to_string()),
+                "FAIL" => Some("Fail".to_string()),
                 _ => None,
             });
-            let behavior_enum = behavior.and_then(|b| match b.to_uppercase().as_str() {
-                "RESTRICT" => Some(drop_namespace_request::Behavior::Restrict),
-                "CASCADE" => Some(drop_namespace_request::Behavior::Cascade),
+            let behavior_str = behavior.and_then(|b| match b.to_uppercase().as_str() {
+                "RESTRICT" => Some("Restrict".to_string()),
+                "CASCADE" => Some("Cascade".to_string()),
                 _ => None,
             });
             let request = DropNamespaceRequest {
@@ -377,8 +381,9 @@ impl Connection {
                 } else {
                     Some(namespace)
                 },
-                mode: mode_enum,
-                behavior: behavior_enum,
+                mode: mode_str,
+                behavior: behavior_str,
+                ..Default::default()
             };
             let response = inner.drop_namespace(request).await.infer_error()?;
             Python::with_gil(|py| -> PyResult<Py<PyDict>> {
@@ -405,6 +410,7 @@ impl Connection {
                 } else {
                     Some(namespace)
                 },
+                ..Default::default()
             };
             let response = inner.describe_namespace(request).await.infer_error()?;
             Python::with_gil(|py| -> PyResult<Py<PyDict>> {
@@ -434,6 +440,7 @@ impl Connection {
                 },
                 page_token,
                 limit: limit.map(|l| l as i32),
+                ..Default::default()
             };
             let response = inner.list_tables(request).await.infer_error()?;
             Python::with_gil(|py| -> PyResult<Py<PyDict>> {

python/src/table.rs

@@ -497,6 +497,11 @@ impl Table {
         })
     }
 
+    pub fn uri(self_: PyRef<'_, Self>) -> PyResult<Bound<'_, PyAny>> {
+        let inner = self_.inner_ref()?.clone();
+        future_into_py(self_.py(), async move { inner.uri().await.infer_error() })
+    }
+
     pub fn __repr__(&self) -> String {
         match &self.inner {
             None => format!("ClosedTable({})", self.name),

rust/lancedb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lancedb"
-version = "0.23.0-beta.0"
+version = "0.24.1"
 edition.workspace = true
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license.workspace = true
@@ -104,13 +104,18 @@ test-log = "0.2"
 
 
 [features]
-default = ["aws", "gcs", "azure", "dynamodb", "oss"]
+default = []
 aws = ["lance/aws", "lance-io/aws", "lance-namespace-impls/dir-aws"]
 oss = ["lance/oss", "lance-io/oss", "lance-namespace-impls/dir-oss"]
 gcs = ["lance/gcp", "lance-io/gcp", "lance-namespace-impls/dir-gcp"]
 azure = ["lance/azure", "lance-io/azure", "lance-namespace-impls/dir-azure"]
+huggingface = [
+    "lance/huggingface",
+    "lance-io/huggingface",
+    "lance-namespace-impls/dir-huggingface",
+]
 dynamodb = ["lance/dynamodb", "aws"]
-remote = ["dep:reqwest", "dep:http", "lance-namespace-impls/rest"]
+remote = ["dep:reqwest", "dep:http", "lance-namespace-impls/rest", "lance-namespace-impls/rest-adapter"]
 fp16kernels = ["lance-linalg/fp16kernels"]
 s3-test = []
 bedrock = ["dep:aws-sdk-bedrockruntime"]
@@ -148,3 +153,6 @@ name = "ivf_pq"
 [[example]]
 name = "hybrid_search"
 required-features = ["sentence-transformers"]
+
+[package.metadata.docs.rs]
+all-features = true

rust/lancedb/src/connection.rs

@@ -36,10 +36,42 @@ use crate::remote::{
 };
 use crate::table::{TableDefinition, WriteOptions};
 use crate::Table;
+use lance::io::ObjectStoreParams;
 pub use lance_encoding::version::LanceFileVersion;
 #[cfg(feature = "remote")]
 use lance_io::object_store::StorageOptions;
-use lance_io::object_store::StorageOptionsProvider;
+use lance_io::object_store::{StorageOptionsAccessor, StorageOptionsProvider};
+
+fn merge_storage_options(
+    store_params: &mut ObjectStoreParams,
+    pairs: impl IntoIterator<Item = (String, String)>,
+) {
+    let mut options = store_params.storage_options().cloned().unwrap_or_default();
+    for (key, value) in pairs {
+        options.insert(key, value);
+    }
+    let provider = store_params
+        .storage_options_accessor
+        .as_ref()
+        .and_then(|accessor| accessor.provider().cloned());
+    let accessor = if let Some(provider) = provider {
+        StorageOptionsAccessor::with_initial_and_provider(options, provider)
+    } else {
+        StorageOptionsAccessor::with_static_options(options)
+    };
+    store_params.storage_options_accessor = Some(Arc::new(accessor));
+}
+
+fn set_storage_options_provider(
+    store_params: &mut ObjectStoreParams,
+    provider: Arc<dyn StorageOptionsProvider>,
+) {
+    let accessor = match store_params.storage_options().cloned() {
+        Some(options) => StorageOptionsAccessor::with_initial_and_provider(options, provider),
+        None => StorageOptionsAccessor::with_provider(provider),
+    };
+    store_params.storage_options_accessor = Some(Arc::new(accessor));
+}
 
 /// A builder for configuring a [`Connection::table_names`] operation
 pub struct TableNamesBuilder {
@@ -219,8 +251,36 @@ impl CreateTableBuilder<false> {
     /// Execute the create table operation
     pub async fn execute(self) -> Result<Table> {
         let parent = self.parent.clone();
-        let table = parent.create_table(self.request).await?;
-        Ok(Table::new(table, parent))
+        let embedding_registry = self.embedding_registry.clone();
+        let request = self.into_request()?;
+        Ok(Table::new_with_embedding_registry(
+            parent.create_table(request).await?,
+            parent,
+            embedding_registry,
+        ))
+    }
+
+    fn into_request(self) -> Result<CreateTableRequest> {
+        if self.embeddings.is_empty() {
+            return Ok(self.request);
+        }
+
+        let CreateTableData::Empty(table_def) = self.request.data else {
+            unreachable!("CreateTableBuilder<false> should always have Empty data")
+        };
+
+        let schema = table_def.schema.clone();
+        let empty_batch = arrow_array::RecordBatch::new_empty(schema.clone());
+
+        let reader = Box::new(std::iter::once(Ok(empty_batch)).collect::<Vec<_>>());
+        let reader = arrow_array::RecordBatchIterator::new(reader.into_iter(), schema);
+        let with_embeddings = WithEmbeddings::new(reader, self.embeddings);
+        let table_definition = with_embeddings.table_definition()?;
+
+        Ok(CreateTableRequest {
+            data: CreateTableData::Empty(table_definition),
+            ..self.request
+        })
     }
 }
 
@@ -246,16 +306,14 @@ impl<const HAS_DATA: bool> CreateTableBuilder<HAS_DATA> {
     ///
     /// See available options at <https://lancedb.com/docs/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
-        let store_options = self
+        let store_params = self
             .request
             .write_options
             .lance_write_params
             .get_or_insert(Default::default())
             .store_params
-            .get_or_insert(Default::default())
-            .storage_options
             .get_or_insert(Default::default());
-        store_options.insert(key.into(), value.into());
+        merge_storage_options(store_params, [(key.into(), value.into())]);
         self
     }
 
@@ -269,19 +327,17 @@ impl<const HAS_DATA: bool> CreateTableBuilder<HAS_DATA> {
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
     ) -> Self {
-        let store_options = self
+        let store_params = self
             .request
             .write_options
             .lance_write_params
             .get_or_insert(Default::default())
             .store_params
-            .get_or_insert(Default::default())
-            .storage_options
             .get_or_insert(Default::default());
-
-        for (key, value) in pairs {
-            store_options.insert(key.into(), value.into());
-        }
+        let updates = pairs
+            .into_iter()
+            .map(|(key, value)| (key.into(), value.into()));
+        merge_storage_options(store_params, updates);
         self
     }
 
@@ -318,23 +374,21 @@ impl<const HAS_DATA: bool> CreateTableBuilder<HAS_DATA> {
     /// This has no effect in LanceDB Cloud.
     #[deprecated(since = "0.15.1", note = "Use `database_options` instead")]
     pub fn enable_v2_manifest_paths(mut self, use_v2_manifest_paths: bool) -> Self {
-        let storage_options = self
+        let store_params = self
             .request
             .write_options
             .lance_write_params
             .get_or_insert_with(Default::default)
             .store_params
-            .get_or_insert_with(Default::default)
-            .storage_options
             .get_or_insert_with(Default::default);
-
-        storage_options.insert(
-            OPT_NEW_TABLE_V2_MANIFEST_PATHS.to_string(),
-            if use_v2_manifest_paths {
-                "true".to_string()
-            } else {
-                "false".to_string()
-            },
+        let value = if use_v2_manifest_paths {
+            "true".to_string()
+        } else {
+            "false".to_string()
+        };
+        merge_storage_options(
+            store_params,
+            [(OPT_NEW_TABLE_V2_MANIFEST_PATHS.to_string(), value)],
         );
         self
     }
@@ -344,19 +398,19 @@ impl<const HAS_DATA: bool> CreateTableBuilder<HAS_DATA> {
     /// The default is `LanceFileVersion::Stable`.
     #[deprecated(since = "0.15.1", note = "Use `database_options` instead")]
     pub fn data_storage_version(mut self, data_storage_version: LanceFileVersion) -> Self {
-        let storage_options = self
+        let store_params = self
             .request
             .write_options
             .lance_write_params
             .get_or_insert_with(Default::default)
             .store_params
-            .get_or_insert_with(Default::default)
-            .storage_options
             .get_or_insert_with(Default::default);
-
-        storage_options.insert(
-            OPT_NEW_TABLE_STORAGE_VERSION.to_string(),
-            data_storage_version.to_string(),
+        merge_storage_options(
+            store_params,
+            [(
+                OPT_NEW_TABLE_STORAGE_VERSION.to_string(),
+                data_storage_version.to_string(),
+            )],
         );
         self
     }
@@ -381,13 +435,14 @@ impl<const HAS_DATA: bool> CreateTableBuilder<HAS_DATA> {
     /// This allows tables to automatically refresh cloud storage credentials
     /// when they expire, enabling long-running operations on remote storage.
     pub fn storage_options_provider(mut self, provider: Arc<dyn StorageOptionsProvider>) -> Self {
-        self.request
+        let store_params = self
+            .request
             .write_options
             .lance_write_params
             .get_or_insert(Default::default())
             .store_params
-            .get_or_insert(Default::default())
-            .storage_options_provider = Some(provider);
+            .get_or_insert(Default::default());
+        set_storage_options_provider(store_params, provider);
         self
     }
 }
@@ -450,15 +505,13 @@ impl OpenTableBuilder {
     ///
     /// See available options at <https://lancedb.com/docs/storage/>
     pub fn storage_option(mut self, key: impl Into<String>, value: impl Into<String>) -> Self {
-        let storage_options = self
+        let store_params = self
             .request
             .lance_read_params
             .get_or_insert(Default::default())
             .store_options
-            .get_or_insert(Default::default())
-            .storage_options
             .get_or_insert(Default::default());
-        storage_options.insert(key.into(), value.into());
+        merge_storage_options(store_params, [(key.into(), value.into())]);
         self
     }
 
@@ -472,18 +525,16 @@ impl OpenTableBuilder {
         mut self,
         pairs: impl IntoIterator<Item = (impl Into<String>, impl Into<String>)>,
     ) -> Self {
-        let storage_options = self
+        let store_params = self
             .request
             .lance_read_params
             .get_or_insert(Default::default())
             .store_options
-            .get_or_insert(Default::default())
-            .storage_options
             .get_or_insert(Default::default());
-
-        for (key, value) in pairs {
-            storage_options.insert(key.into(), value.into());
-        }
+        let updates = pairs
+            .into_iter()
+            .map(|(key, value)| (key.into(), value.into()));
+        merge_storage_options(store_params, updates);
         self
     }
 
@@ -507,12 +558,13 @@ impl OpenTableBuilder {
     /// This allows tables to automatically refresh cloud storage credentials
     /// when they expire, enabling long-running operations on remote storage.
     pub fn storage_options_provider(mut self, provider: Arc<dyn StorageOptionsProvider>) -> Self {
-        self.request
+        let store_params = self
+            .request
             .lance_read_params
             .get_or_insert(Default::default())
             .store_options
-            .get_or_insert(Default::default())
-            .storage_options_provider = Some(provider);
+            .get_or_insert(Default::default());
+        set_storage_options_provider(store_params, provider);
         self
     }
 
@@ -804,6 +856,14 @@ impl Connection {
         self.internal.describe_namespace(request).await
     }
 
+    /// Get the equivalent namespace client in the database of this connection.
+    /// For LanceNamespaceDatabase, it is the underlying LanceNamespace.
+    /// For ListingDatabase, it is the equivalent DirectoryNamespace.
+    /// For RemoteDatabase, it is the equivalent RestNamespace.
+    pub async fn namespace_client(&self) -> Result<Arc<dyn lance_namespace::LanceNamespace>> {
+        self.internal.namespace_client().await
+    }
+
     /// List tables with pagination support
     pub async fn list_tables(&self, request: ListTablesRequest) -> Result<ListTablesResponse> {
         self.internal.list_tables(request).await
@@ -860,6 +920,10 @@ pub struct ConnectBuilder {
     embedding_registry: Option<Arc<dyn EmbeddingRegistry>>,
 }
 
+#[cfg(feature = "remote")]
+const ENV_VARS_TO_STORAGE_OPTS: [(&str, &str); 1] =
+    [("AZURE_STORAGE_ACCOUNT_NAME", "azure_storage_account_name")];
+
 impl ConnectBuilder {
     /// Create a new [`ConnectOptions`] with the given database URI.
     pub fn new(uri: &str) -> Self {
@@ -1043,11 +1107,27 @@ impl ConnectBuilder {
         self
     }
 
+    #[cfg(feature = "remote")]
+    fn apply_env_defaults(
+        env_var_to_remote_storage_option: &[(&str, &str)],
+        options: &mut HashMap<String, String>,
+    ) {
+        for (env_key, opt_key) in env_var_to_remote_storage_option {
+            if let Ok(env_value) = std::env::var(env_key) {
+                if !options.contains_key(*opt_key) {
+                    options.insert((*opt_key).to_string(), env_value);
+                }
+            }
+        }
+    }
+
     #[cfg(feature = "remote")]
     fn execute_remote(self) -> Result<Connection> {
         use crate::remote::db::RemoteDatabaseOptions;
 
-        let options = RemoteDatabaseOptions::parse_from_map(&self.request.options)?;
+        let mut merged_options = self.request.options.clone();
+        Self::apply_env_defaults(&ENV_VARS_TO_STORAGE_OPTS, &mut merged_options);
+        let options = RemoteDatabaseOptions::parse_from_map(&merged_options)?;
 
         let region = options.region.ok_or_else(|| Error::InvalidInput {
             message: "A region is required when connecting to LanceDb Cloud".to_string(),
@@ -1269,8 +1349,6 @@ mod test_utils {
 
 #[cfg(test)]
 mod tests {
-    use std::fs::create_dir_all;
-
     use crate::database::listing::{ListingDatabaseOptions, NewTableConfig};
     use crate::query::QueryBase;
     use crate::query::{ExecutableQuery, QueryExecutionOptions};
@@ -1294,6 +1372,23 @@ mod tests {
         assert_eq!(tc.connection.uri(), tc.uri);
     }
 
+    #[cfg(feature = "remote")]
+    #[test]
+    fn test_apply_env_defaults() {
+        let env_key = "TEST_APPLY_ENV_DEFAULTS_ENVIRONMENT_VARIABLE_ENV_KEY";
+        let env_val = "TEST_APPLY_ENV_DEFAULTS_ENVIRONMENT_VARIABLE_ENV_VAL";
+        let opts_key = "test_apply_env_defaults_environment_variable_opts_key";
+        std::env::set_var(env_key, env_val);
+
+        let mut options = HashMap::new();
+        ConnectBuilder::apply_env_defaults(&[(env_key, opts_key)], &mut options);
+        assert_eq!(Some(&env_val.to_string()), options.get(opts_key));
+
+        options.insert(opts_key.to_string(), "EXPLICIT-VALUE".to_string());
+        ConnectBuilder::apply_env_defaults(&[(env_key, opts_key)], &mut options);
+        assert_eq!(Some(&"EXPLICIT-VALUE".to_string()), options.get(opts_key));
+    }
+
     #[cfg(not(windows))]
     #[tokio::test]
     async fn test_connect_relative() {
@@ -1317,25 +1412,27 @@ mod tests {
 
     #[tokio::test]
     async fn test_table_names() {
-        let tmp_dir = tempdir().unwrap();
+        let tc = new_test_connection().await.unwrap();
+        let db = tc.connection;
+        let schema = Arc::new(Schema::new(vec![Field::new("x", DataType::Int32, false)]));
         let mut names = Vec::with_capacity(100);
         for _ in 0..100 {
-            let mut name = uuid::Uuid::new_v4().to_string();
+            let name = uuid::Uuid::new_v4().to_string();
             names.push(name.clone());
-            name.push_str(".lance");
-            create_dir_all(tmp_dir.path().join(&name)).unwrap();
+            db.create_empty_table(name, schema.clone())
+                .execute()
+                .await
+                .unwrap();
         }
         names.sort();
-
-        let uri = tmp_dir.path().to_str().unwrap();
-        let db = connect(uri).execute().await.unwrap();
-        let tables = db.table_names().execute().await.unwrap();
+        let tables = db.table_names().limit(100).execute().await.unwrap();
 
         assert_eq!(tables, names);
 
         let tables = db
             .table_names()
             .start_after(&names[30])
+            .limit(100)
             .execute()
             .await
             .unwrap();
@@ -1516,18 +1613,27 @@ mod tests {
 
     #[tokio::test]
     async fn drop_table() {
-        let tmp_dir = tempdir().unwrap();
+        let tc = new_test_connection().await.unwrap();
+        let db = tc.connection;
 
-        let uri = tmp_dir.path().to_str().unwrap();
-        let db = connect(uri).execute().await.unwrap();
+        if tc.is_remote {
+            // All the typical endpoints such as s3:///, file-object-store:///, etc. treat drop_table
+            // as idempotent.
+            assert!(db.drop_table("invalid_table", &[]).await.is_ok());
+        } else {
+            // The behavior of drop_table when using a file:/// endpoint differs from all other
+            // object providers, in that it returns an error when deleting a non-existent table.
+            assert!(matches!(
+                db.drop_table("invalid_table", &[]).await,
+                Err(crate::Error::TableNotFound { .. }),
+            ));
+        }
 
-        // drop non-exist table
-        assert!(matches!(
-            db.drop_table("invalid_table", &[]).await,
-            Err(crate::Error::TableNotFound { .. }),
-        ));
-
-        create_dir_all(tmp_dir.path().join("table1.lance")).unwrap();
+        let schema = Arc::new(Schema::new(vec![Field::new("x", DataType::Int32, false)]));
+        db.create_empty_table("table1", schema.clone())
+            .execute()
+            .await
+            .unwrap();
         db.drop_table("table1", &[]).await.unwrap();
 
         let tables = db.table_names().execute().await.unwrap();
@@ -1614,4 +1720,128 @@ mod tests {
         let cloned_count = cloned_table.count_rows(None).await.unwrap();
         assert_eq!(source_count, cloned_count);
     }
+
+    #[tokio::test]
+    async fn test_create_empty_table_with_embeddings() {
+        use crate::embeddings::{EmbeddingDefinition, EmbeddingFunction};
+        use arrow_array::{
+            Array, FixedSizeListArray, Float32Array, RecordBatch, RecordBatchIterator, StringArray,
+        };
+        use std::borrow::Cow;
+
+        #[derive(Debug, Clone)]
+        struct MockEmbedding {
+            dim: usize,
+        }
+
+        impl EmbeddingFunction for MockEmbedding {
+            fn name(&self) -> &str {
+                "test_embedding"
+            }
+
+            fn source_type(&self) -> Result<Cow<'_, DataType>> {
+                Ok(Cow::Owned(DataType::Utf8))
+            }
+
+            fn dest_type(&self) -> Result<Cow<'_, DataType>> {
+                Ok(Cow::Owned(DataType::new_fixed_size_list(
+                    DataType::Float32,
+                    self.dim as i32,
+                    true,
+                )))
+            }
+
+            fn compute_source_embeddings(&self, source: Arc<dyn Array>) -> Result<Arc<dyn Array>> {
+                let len = source.len();
+                let values = vec![1.0f32; len * self.dim];
+                let values = Arc::new(Float32Array::from(values));
+                let field = Arc::new(Field::new("item", DataType::Float32, true));
+                Ok(Arc::new(FixedSizeListArray::new(
+                    field,
+                    self.dim as i32,
+                    values,
+                    None,
+                )))
+            }
+
+            fn compute_query_embeddings(&self, _input: Arc<dyn Array>) -> Result<Arc<dyn Array>> {
+                unimplemented!()
+            }
+        }
+
+        let tmp_dir = tempdir().unwrap();
+        let uri = tmp_dir.path().to_str().unwrap();
+        let db = connect(uri).execute().await.unwrap();
+
+        let embed_func = Arc::new(MockEmbedding { dim: 128 });
+        db.embedding_registry()
+            .register("test_embedding", embed_func.clone())
+            .unwrap();
+
+        let schema = Arc::new(Schema::new(vec![Field::new("name", DataType::Utf8, true)]));
+        let ed = EmbeddingDefinition {
+            source_column: "name".to_owned(),
+            dest_column: Some("name_embedding".to_owned()),
+            embedding_name: "test_embedding".to_owned(),
+        };
+
+        let table = db
+            .create_empty_table("test", schema)
+            .mode(CreateTableMode::Overwrite)
+            .add_embedding(ed)
+            .unwrap()
+            .execute()
+            .await
+            .unwrap();
+
+        let table_schema = table.schema().await.unwrap();
+        assert!(table_schema.column_with_name("name").is_some());
+        assert!(table_schema.column_with_name("name_embedding").is_some());
+
+        let embedding_field = table_schema.field_with_name("name_embedding").unwrap();
+        assert_eq!(
+            embedding_field.data_type(),
+            &DataType::new_fixed_size_list(DataType::Float32, 128, true)
+        );
+
+        let input_schema = Arc::new(Schema::new(vec![Field::new("name", DataType::Utf8, true)]));
+        let input_batch = RecordBatch::try_new(
+            input_schema.clone(),
+            vec![Arc::new(StringArray::from(vec![
+                Some("Alice"),
+                Some("Bob"),
+                Some("Charlie"),
+            ]))],
+        )
+        .unwrap();
+
+        let input_reader = Box::new(RecordBatchIterator::new(
+            vec![Ok(input_batch)].into_iter(),
+            input_schema,
+        ));
+
+        table.add(input_reader).execute().await.unwrap();
+
+        let results = table
+            .query()
+            .execute()
+            .await
+            .unwrap()
+            .try_collect::<Vec<_>>()
+            .await
+            .unwrap();
+
+        assert_eq!(results.len(), 1);
+        let batch = &results[0];
+        assert_eq!(batch.num_rows(), 3);
+        assert!(batch.column_by_name("name_embedding").is_some());
+
+        let embedding_col = batch
+            .column_by_name("name_embedding")
+            .unwrap()
+            .as_any()
+            .downcast_ref::<FixedSizeListArray>()
+            .unwrap();
+        assert_eq!(embedding_col.len(), 3);
+    }
 }

rust/lancedb/src/database.rs

@@ -296,4 +296,10 @@ pub trait Database:
     /// Drop all tables in the database
     async fn drop_all_tables(&self, namespace: &[String]) -> Result<()>;
     fn as_any(&self) -> &dyn std::any::Any;
+
+    /// Get the equivalent namespace client of this database
+    /// For LanceNamespaceDatabase, it is the underlying LanceNamespace.
+    /// For ListingDatabase, it is the equivalent DirectoryNamespace.
+    /// For RemoteDatabase, it is the equivalent RestNamespace.
+    async fn namespace_client(&self) -> Result<Arc<dyn LanceNamespace>>;
 }

rust/lancedb/src/database/listing.rs

@@ -12,7 +12,7 @@ use lance::dataset::{builder::DatasetBuilder, ReadParams, WriteMode};
 use lance::io::{ObjectStore, ObjectStoreParams, WrappingObjectStore};
 use lance_datafusion::utils::StreamingWriteSource;
 use lance_encoding::version::LanceFileVersion;
-use lance_io::object_store::StorageOptionsProvider;
+use lance_io::object_store::{StorageOptionsAccessor, StorageOptionsProvider};
 use lance_table::io::commit::commit_handler_from_url;
 use object_store::local::LocalFileSystem;
 use snafu::ResultExt;
@@ -356,7 +356,13 @@ impl ListingDatabase {
                     .clone()
                     .unwrap_or_else(|| Arc::new(lance::session::Session::default()));
                 let os_params = ObjectStoreParams {
-                    storage_options: Some(options.storage_options.clone()),
+                    storage_options_accessor: if options.storage_options.is_empty() {
+                        None
+                    } else {
+                        Some(Arc::new(StorageOptionsAccessor::with_static_options(
+                            options.storage_options.clone(),
+                        )))
+                    },
                     ..Default::default()
                 };
                 let (object_store, base_path) = ObjectStore::from_uri_and_params(
@@ -463,9 +469,20 @@ impl ListingDatabase {
         validate_table_name(name)?;
 
         let mut uri = self.uri.clone();
-        // If the URI does not end with a slash, add one
-        if !uri.ends_with('/') {
-            uri.push('/');
+        // If the URI does not end with a path separator, add one
+        // Use forward slash for URIs (http://, s3://, gs://, file://, etc.)
+        // Use platform-specific separator for local paths without scheme
+        let has_scheme = uri.contains("://");
+        let ends_with_separator = uri.ends_with('/') || uri.ends_with('\\');
+
+        if !ends_with_separator {
+            if has_scheme {
+                // URIs always use forward slash
+                uri.push('/');
+            } else {
+                // Local path without scheme - use platform separator
+                uri.push(std::path::MAIN_SEPARATOR);
+            }
         }
         // Append the table name with the lance file extension
         uri.push_str(&format!("{}.{}", name, LANCE_FILE_EXTENSION));
@@ -481,7 +498,13 @@ impl ListingDatabase {
 
     async fn drop_tables(&self, names: Vec<String>) -> Result<()> {
         let object_store_params = ObjectStoreParams {
-            storage_options: Some(self.storage_options.clone()),
+            storage_options_accessor: if self.storage_options.is_empty() {
+                None
+            } else {
+                Some(Arc::new(StorageOptionsAccessor::with_static_options(
+                    self.storage_options.clone(),
+                )))
+            },
             ..Default::default()
         };
         let mut uri = self.uri.clone();
@@ -530,7 +553,7 @@ impl ListingDatabase {
             .lance_write_params
             .as_ref()
             .and_then(|p| p.store_params.as_ref())
-            .and_then(|sp| sp.storage_options.as_ref());
+            .and_then(|sp| sp.storage_options());
 
         let storage_version_override = storage_options
             .and_then(|opts| opts.get(OPT_NEW_TABLE_STORAGE_VERSION))
@@ -581,21 +604,20 @@ impl ListingDatabase {
         // will cause a new connection to be created, and that connection will
         // be dropped from the cache when python GCs the table object, which
         // confounds reuse across tables.
-        if !self.storage_options.is_empty() {
-            let storage_options = write_params
+        if !self.storage_options.is_empty() || self.storage_options_provider.is_some() {
+            let store_params = write_params
                 .store_params
-                .get_or_insert_with(Default::default)
-                .storage_options
                 .get_or_insert_with(Default::default);
-            self.inherit_storage_options(storage_options);
-        }
-
-        // Set storage options provider if available
-        if self.storage_options_provider.is_some() {
-            write_params
-                .store_params
-                .get_or_insert_with(Default::default)
-                .storage_options_provider = self.storage_options_provider.clone();
+            let mut storage_options = store_params.storage_options().cloned().unwrap_or_default();
+            if !self.storage_options.is_empty() {
+                self.inherit_storage_options(&mut storage_options);
+            }
+            let accessor = if let Some(ref provider) = self.storage_options_provider {
+                StorageOptionsAccessor::with_initial_and_provider(storage_options, provider.clone())
+            } else {
+                StorageOptionsAccessor::with_static_options(storage_options)
+            };
+            store_params.storage_options_accessor = Some(Arc::new(accessor));
         }
 
         write_params.data_storage_version = self
@@ -881,7 +903,13 @@ impl Database for ListingDatabase {
         validate_table_name(&request.target_table_name)?;
 
         let storage_params = ObjectStoreParams {
-            storage_options: Some(self.storage_options.clone()),
+            storage_options_accessor: if self.storage_options.is_empty() {
+                None
+            } else {
+                Some(Arc::new(StorageOptionsAccessor::with_static_options(
+                    self.storage_options.clone(),
+                )))
+            },
             ..Default::default()
         };
         let read_params = ReadParams {
@@ -945,25 +973,28 @@ impl Database for ListingDatabase {
         // will cause a new connection to be created, and that connection will
         // be dropped from the cache when python GCs the table object, which
         // confounds reuse across tables.
-        if !self.storage_options.is_empty() {
-            let storage_options = request
+        if !self.storage_options.is_empty() || self.storage_options_provider.is_some() {
+            let store_params = request
                 .lance_read_params
                 .get_or_insert_with(Default::default)
                 .store_options
-                .get_or_insert_with(Default::default)
-                .storage_options
                 .get_or_insert_with(Default::default);
-            self.inherit_storage_options(storage_options);
-        }
-
-        // Set storage options provider if available
-        if self.storage_options_provider.is_some() {
-            request
-                .lance_read_params
-                .get_or_insert_with(Default::default)
-                .store_options
-                .get_or_insert_with(Default::default)
-                .storage_options_provider = self.storage_options_provider.clone();
+            let mut storage_options = store_params.storage_options().cloned().unwrap_or_default();
+            if !self.storage_options.is_empty() {
+                self.inherit_storage_options(&mut storage_options);
+            }
+            // Preserve request-level provider if no connection-level provider exists
+            let request_provider = store_params
+                .storage_options_accessor
+                .as_ref()
+                .and_then(|a| a.provider().cloned());
+            let provider = self.storage_options_provider.clone().or(request_provider);
+            let accessor = if let Some(provider) = provider {
+                StorageOptionsAccessor::with_initial_and_provider(storage_options, provider)
+            } else {
+                StorageOptionsAccessor::with_static_options(storage_options)
+            };
+            store_params.storage_options_accessor = Some(Arc::new(accessor));
         }
 
         // Some ReadParams are exposed in the OpenTableBuilder, but we also
@@ -1043,6 +1074,24 @@ impl Database for ListingDatabase {
     fn as_any(&self) -> &dyn std::any::Any {
         self
     }
+
+    async fn namespace_client(&self) -> Result<Arc<dyn lance_namespace::LanceNamespace>> {
+        // Create a DirectoryNamespace pointing to the same root with the same storage options
+        let mut builder = lance_namespace_impls::DirectoryNamespaceBuilder::new(&self.uri);
+
+        // Add storage options
+        if !self.storage_options.is_empty() {
+            builder = builder.storage_options(self.storage_options.clone());
+        }
+
+        // Use the same session
+        builder = builder.session(self.session.clone());
+
+        let namespace = builder.build().await.map_err(|e| Error::Runtime {
+            message: format!("Failed to create namespace client: {}", e),
+        })?;
+        Ok(Arc::new(namespace) as Arc<dyn lance_namespace::LanceNamespace>)
+    }
 }
 
 #[cfg(test)]
@@ -1053,6 +1102,7 @@ mod tests {
     use crate::table::{Table, TableDefinition};
     use arrow_array::{Int32Array, RecordBatch, StringArray};
     use arrow_schema::{DataType, Field, Schema};
+    use std::path::PathBuf;
     use tempfile::tempdir;
 
     async fn setup_database() -> (tempfile::TempDir, ListingDatabase) {
@@ -1851,7 +1901,9 @@ mod tests {
         let write_options = WriteOptions {
             lance_write_params: Some(lance::dataset::WriteParams {
                 store_params: Some(lance::io::ObjectStoreParams {
-                    storage_options: Some(storage_options),
+                    storage_options_accessor: Some(Arc::new(
+                        StorageOptionsAccessor::with_static_options(storage_options),
+                    )),
                     ..Default::default()
                 }),
                 ..Default::default()
@@ -1925,7 +1977,9 @@ mod tests {
         let write_options = WriteOptions {
             lance_write_params: Some(lance::dataset::WriteParams {
                 store_params: Some(lance::io::ObjectStoreParams {
-                    storage_options: Some(storage_options),
+                    storage_options_accessor: Some(Arc::new(
+                        StorageOptionsAccessor::with_static_options(storage_options),
+                    )),
                     ..Default::default()
                 }),
                 ..Default::default()
@@ -2027,4 +2081,76 @@ mod tests {
         let db_options = ListingDatabaseOptions::parse_from_map(&options).unwrap();
         assert_eq!(db_options.new_table_config.enable_stable_row_ids, None);
     }
+
+    #[tokio::test]
+    async fn test_table_uri() {
+        let (_tempdir, db) = setup_database().await;
+
+        let mut pb = PathBuf::new();
+        pb.push(db.uri.clone());
+        pb.push("test.lance");
+
+        let expected = pb.to_str().unwrap();
+        let uri = db.table_uri("test").ok().unwrap();
+        assert_eq!(uri, expected);
+    }
+
+    #[tokio::test]
+    async fn test_namespace_client() {
+        let (_tempdir, db) = setup_database().await;
+
+        // Create some tables first
+        let schema = Arc::new(Schema::new(vec![
+            Field::new("id", DataType::Int32, false),
+            Field::new("name", DataType::Utf8, false),
+        ]));
+
+        db.create_table(CreateTableRequest {
+            name: "table1".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema.clone())),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+            location: None,
+            namespace_client: None,
+        })
+        .await
+        .unwrap();
+
+        db.create_table(CreateTableRequest {
+            name: "table2".to_string(),
+            namespace: vec![],
+            data: CreateTableData::Empty(TableDefinition::new_from_schema(schema)),
+            mode: CreateTableMode::Create,
+            write_options: Default::default(),
+            location: None,
+            namespace_client: None,
+        })
+        .await
+        .unwrap();
+
+        // Get the namespace client
+        let namespace_client = db.namespace_client().await;
+        assert!(namespace_client.is_ok());
+        let namespace_client = namespace_client.unwrap();
+
+        // Verify the namespace client can list the tables we created
+        // Use empty vec for root namespace
+        let list_result = namespace_client
+            .list_tables(lance_namespace::models::ListTablesRequest {
+                id: Some(vec![]),
+                ..Default::default()
+            })
+            .await;
+        assert!(
+            list_result.is_ok(),
+            "list_tables failed: {:?}",
+            list_result.err()
+        );
+
+        let tables = list_result.unwrap().tables;
+        assert_eq!(tables.len(), 2);
+        assert!(tables.contains(&"table1".to_string()));
+        assert!(tables.contains(&"table2".to_string()));
+    }
 }

rust/lancedb/src/database/namespace.rs

@@ -7,25 +7,25 @@ use std::collections::HashMap;
 use std::sync::Arc;
 
 use async_trait::async_trait;
-use lance_io::object_store::{LanceNamespaceStorageOptionsProvider, StorageOptionsProvider};
 use lance_namespace::{
     models::{
         CreateEmptyTableRequest, CreateNamespaceRequest, CreateNamespaceResponse,
-        DescribeNamespaceRequest, DescribeNamespaceResponse, DescribeTableRequest,
-        DropNamespaceRequest, DropNamespaceResponse, DropTableRequest, ListNamespacesRequest,
-        ListNamespacesResponse, ListTablesRequest, ListTablesResponse,
+        DeclareTableRequest, DescribeNamespaceRequest, DescribeNamespaceResponse,
+        DescribeTableRequest, DropNamespaceRequest, DropNamespaceResponse, DropTableRequest,
+        ListNamespacesRequest, ListNamespacesResponse, ListTablesRequest, ListTablesResponse,
     },
     LanceNamespace,
 };
 use lance_namespace_impls::ConnectBuilder;
+use log::warn;
 
-use crate::connection::ConnectRequest;
 use crate::database::ReadConsistency;
 use crate::error::{Error, Result};
+use crate::table::NativeTable;
 
 use super::{
-    listing::ListingDatabase, BaseTable, CloneTableRequest, CreateTableMode,
-    CreateTableRequest as DbCreateTableRequest, Database, OpenTableRequest, TableNamesRequest,
+    BaseTable, CloneTableRequest, CreateTableMode, CreateTableRequest as DbCreateTableRequest,
+    Database, OpenTableRequest, TableNamesRequest,
 };
 
 /// A database implementation that uses lance-namespace for table management
@@ -90,51 +90,6 @@ impl std::fmt::Display for LanceNamespaceDatabase {
     }
 }
 
-impl LanceNamespaceDatabase {
-    /// Create a temporary listing database for the given location
-    ///
-    /// Merges storage options with priority: connection < user < namespace
-    async fn create_listing_database(
-        &self,
-        location: &str,
-        table_id: Vec<String>,
-        user_storage_options: Option<&HashMap<String, String>>,
-        response_storage_options: Option<&HashMap<String, String>>,
-    ) -> Result<ListingDatabase> {
-        // Merge storage options: connection < user < namespace
-        let mut merged_storage_options = self.storage_options.clone();
-        if let Some(opts) = user_storage_options {
-            merged_storage_options.extend(opts.clone());
-        }
-        if let Some(opts) = response_storage_options {
-            merged_storage_options.extend(opts.clone());
-        }
-
-        let request = ConnectRequest {
-            uri: location.to_string(),
-            #[cfg(feature = "remote")]
-            client_config: Default::default(),
-            options: merged_storage_options,
-            read_consistency_interval: self.read_consistency_interval,
-            session: self.session.clone(),
-        };
-
-        let mut listing_db = ListingDatabase::connect_with_options(&request).await?;
-
-        // Create storage options provider only if namespace returned storage options
-        // (not just user-provided options)
-        if response_storage_options.is_some() {
-            let provider = Arc::new(LanceNamespaceStorageOptionsProvider::new(
-                self.namespace.clone(),
-                table_id,
-            )) as Arc<dyn StorageOptionsProvider>;
-            listing_db.storage_options_provider = Some(provider);
-        }
-
-        Ok(listing_db)
-    }
-}
-
 #[async_trait]
 impl Database for LanceNamespaceDatabase {
     fn uri(&self) -> &str {
@@ -183,6 +138,7 @@ impl Database for LanceNamespaceDatabase {
             id: Some(request.namespace),
             page_token: request.start_after,
             limit: request.limit.map(|l| l as i32),
+            ..Default::default()
         };
 
         let response = self.namespace.list_tables(ns_request).await?;
@@ -195,19 +151,11 @@ impl Database for LanceNamespaceDatabase {
     }
 
     async fn create_table(&self, request: DbCreateTableRequest) -> Result<Arc<dyn BaseTable>> {
-        // Extract user-provided storage options from request
-        let user_storage_options = request
-            .write_options
-            .lance_write_params
-            .as_ref()
-            .and_then(|lwp| lwp.store_params.as_ref())
-            .and_then(|sp| sp.storage_options.as_ref());
-
         let mut table_id = request.namespace.clone();
         table_id.push(request.name.clone());
         let describe_request = DescribeTableRequest {
             id: Some(table_id.clone()),
-            version: None,
+            ..Default::default()
         };
 
         let describe_result = self.namespace.describe_table(describe_request).await;
@@ -225,6 +173,7 @@ impl Database for LanceNamespaceDatabase {
                     // Drop the existing table - must succeed
                     let drop_request = DropTableRequest {
                         id: Some(table_id.clone()),
+                        ..Default::default()
                     };
                     self.namespace
                         .drop_table(drop_request)
@@ -235,34 +184,20 @@ impl Database for LanceNamespaceDatabase {
                 }
             }
             CreateTableMode::ExistOk(_) => {
-                if let Ok(response) = describe_result {
-                    let location = response.location.ok_or_else(|| Error::Runtime {
-                        message: "Table location is missing from namespace response".to_string(),
-                    })?;
+                if describe_result.is_ok() {
+                    let native_table = NativeTable::open_from_namespace(
+                        self.namespace.clone(),
+                        &request.name,
+                        request.namespace.clone(),
+                        None,
+                        None,
+                        self.read_consistency_interval,
+                        self.server_side_query_enabled,
+                        self.session.clone(),
+                    )
+                    .await?;
 
-                    let listing_db = self
-                        .create_listing_database(
-                            &location,
-                            table_id.clone(),
-                            user_storage_options,
-                            response.storage_options.as_ref(),
-                        )
-                        .await?;
-
-                    let namespace_client = self
-                        .server_side_query_enabled
-                        .then(|| self.namespace.clone());
-
-                    return listing_db
-                        .open_table(OpenTableRequest {
-                            name: request.name.clone(),
-                            namespace: request.namespace.clone(),
-                            index_cache_size: None,
-                            lance_read_params: None,
-                            location: Some(location),
-                            namespace_client,
-                        })
-                        .await;
+                    return Ok(Arc::new(native_table));
                 }
             }
         }
@@ -270,106 +205,85 @@ impl Database for LanceNamespaceDatabase {
         let mut table_id = request.namespace.clone();
         table_id.push(request.name.clone());
 
-        let create_empty_request = CreateEmptyTableRequest {
+        // Try declare_table first, falling back to create_empty_table for backwards
+        // compatibility with older namespace clients that don't support declare_table
+        let declare_request = DeclareTableRequest {
             id: Some(table_id.clone()),
-            location: None,
-            properties: if self.storage_options.is_empty() {
-                None
-            } else {
-                Some(self.storage_options.clone())
-            },
+            ..Default::default()
         };
 
-        let create_empty_response = self
-            .namespace
-            .create_empty_table(create_empty_request)
-            .await
-            .map_err(|e| Error::Runtime {
-                message: format!("Failed to create empty table: {}", e),
-            })?;
+        let location = match self.namespace.declare_table(declare_request).await {
+            Ok(response) => response.location.ok_or_else(|| Error::Runtime {
+                message: "Table location is missing from declare_table response".to_string(),
+            })?,
+            Err(e) => {
+                // Check if the error is "not supported" and try create_empty_table as fallback
+                let err_str = e.to_string().to_lowercase();
+                if err_str.contains("not supported") || err_str.contains("not implemented") {
+                    warn!(
+                        "declare_table is not supported by the namespace client, \
+                        falling back to deprecated create_empty_table. \
+                        create_empty_table is deprecated and will be removed in Lance 3.0.0. \
+                        Please upgrade your namespace client to support declare_table."
+                    );
+                    #[allow(deprecated)]
+                    let create_empty_request = CreateEmptyTableRequest {
+                        id: Some(table_id.clone()),
+                        ..Default::default()
+                    };
 
-        let location = create_empty_response
-            .location
-            .ok_or_else(|| Error::Runtime {
-                message: "Table location is missing from create_empty_table response".to_string(),
-            })?;
+                    #[allow(deprecated)]
+                    let create_response = self
+                        .namespace
+                        .create_empty_table(create_empty_request)
+                        .await
+                        .map_err(|e| Error::Runtime {
+                            message: format!("Failed to create empty table: {}", e),
+                        })?;
 
-        let listing_db = self
-            .create_listing_database(
-                &location,
-                table_id.clone(),
-                user_storage_options,
-                create_empty_response.storage_options.as_ref(),
-            )
-            .await?;
-
-        let namespace_client = self
-            .server_side_query_enabled
-            .then(|| self.namespace.clone());
-
-        let create_request = DbCreateTableRequest {
-            name: request.name,
-            namespace: request.namespace,
-            data: request.data,
-            mode: request.mode,
-            write_options: request.write_options,
-            location: Some(location),
-            namespace_client,
+                    create_response.location.ok_or_else(|| Error::Runtime {
+                        message: "Table location is missing from create_empty_table response"
+                            .to_string(),
+                    })?
+                } else {
+                    return Err(Error::Runtime {
+                        message: format!("Failed to declare table: {}", e),
+                    });
+                }
+            }
         };
 
-        listing_db.create_table(create_request).await
+        let native_table = NativeTable::create_from_namespace(
+            self.namespace.clone(),
+            &location,
+            &request.name,
+            request.namespace.clone(),
+            request.data,
+            None, // write_store_wrapper not used for namespace connections
+            request.write_options.lance_write_params,
+            self.read_consistency_interval,
+            self.server_side_query_enabled,
+            self.session.clone(),
+        )
+        .await?;
+
+        Ok(Arc::new(native_table))
     }
 
     async fn open_table(&self, request: OpenTableRequest) -> Result<Arc<dyn BaseTable>> {
-        // Extract user-provided storage options from request
-        let user_storage_options = request
-            .lance_read_params
-            .as_ref()
-            .and_then(|lrp| lrp.store_options.as_ref())
-            .and_then(|so| so.storage_options.as_ref());
+        let native_table = NativeTable::open_from_namespace(
+            self.namespace.clone(),
+            &request.name,
+            request.namespace.clone(),
+            None, // write_store_wrapper not used for namespace connections
+            request.lance_read_params,
+            self.read_consistency_interval,
+            self.server_side_query_enabled,
+            self.session.clone(),
+        )
+        .await?;
 
-        let mut table_id = request.namespace.clone();
-        table_id.push(request.name.clone());
-
-        let describe_request = DescribeTableRequest {
-            id: Some(table_id.clone()),
-            version: None,
-        };
-        let response = self
-            .namespace
-            .describe_table(describe_request)
-            .await
-            .map_err(|e| Error::Runtime {
-                message: format!("Failed to describe table: {}", e),
-            })?;
-
-        let location = response.location.ok_or_else(|| Error::Runtime {
-            message: "Table location is missing from namespace response".to_string(),
-        })?;
-
-        let listing_db = self
-            .create_listing_database(
-                &location,
-                table_id.clone(),
-                user_storage_options,
-                response.storage_options.as_ref(),
-            )
-            .await?;
-
-        let namespace_client = self
-            .server_side_query_enabled
-            .then(|| self.namespace.clone());
-
-        let open_request = OpenTableRequest {
-            name: request.name.clone(),
-            namespace: request.namespace.clone(),
-            index_cache_size: request.index_cache_size,
-            lance_read_params: request.lance_read_params,
-            location: Some(location),
-            namespace_client,
-        };
-
-        listing_db.open_table(open_request).await
+        Ok(Arc::new(native_table))
     }
 
     async fn clone_table(&self, _request: CloneTableRequest) -> Result<Arc<dyn BaseTable>> {
@@ -394,7 +308,10 @@ impl Database for LanceNamespaceDatabase {
         let mut table_id = namespace.to_vec();
         table_id.push(name.to_string());
 
-        let drop_request = DropTableRequest { id: Some(table_id) };
+        let drop_request = DropTableRequest {
+            id: Some(table_id),
+            ..Default::default()
+        };
         self.namespace
             .drop_table(drop_request)
             .await
@@ -425,6 +342,10 @@ impl Database for LanceNamespaceDatabase {
     fn as_any(&self) -> &dyn std::any::Any {
         self
     }
+
+    async fn namespace_client(&self) -> Result<Arc<dyn LanceNamespace>> {
+        Ok(self.namespace.clone())
+    }
 }
 
 #[cfg(test)]
@@ -545,8 +466,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");
@@ -606,8 +526,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");
@@ -670,8 +589,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");
@@ -754,8 +672,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");
@@ -810,8 +727,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");
@@ -891,8 +807,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");
@@ -925,8 +840,7 @@ mod tests {
         // Create a child namespace first
         conn.create_namespace(CreateNamespaceRequest {
             id: Some(vec!["test_ns".into()]),
-            mode: None,
-            properties: None,
+            ..Default::default()
         })
         .await
         .expect("Failed to create namespace");

rust/lancedb/src/dataloader/permutation/builder.rs

@@ -19,7 +19,7 @@ use crate::{
         split::{SplitStrategy, Splitter, SPLIT_ID_COLUMN},
         util::{rename_column, TemporaryDirectory},
     },
-    query::{ExecutableQuery, QueryBase},
+    query::{ExecutableQuery, QueryBase, Select},
     Error, Result, Table,
 };
 
@@ -27,6 +27,8 @@ pub const SRC_ROW_ID_COL: &str = "row_id";
 
 pub const SPLIT_NAMES_CONFIG_KEY: &str = "split_names";
 
+pub const DEFAULT_MEMORY_LIMIT: usize = 100 * 1024 * 1024;
+
 /// Where to store the permutation table
 #[derive(Debug, Clone, Default)]
 enum PermutationDestination {
@@ -167,10 +169,20 @@ impl PermutationBuilder {
         &self,
         data: SendableRecordBatchStream,
     ) -> Result<SendableRecordBatchStream> {
+        let memory_limit = std::env::var("LANCEDB_PERM_BUILDER_MEMORY_LIMIT")
+            .unwrap_or_else(|_| DEFAULT_MEMORY_LIMIT.to_string())
+            .parse::<usize>()
+            .unwrap_or_else(|_| {
+                log::error!(
+                    "Failed to parse LANCEDB_PERM_BUILDER_MEMORY_LIMIT, using default: {}",
+                    DEFAULT_MEMORY_LIMIT
+                );
+                DEFAULT_MEMORY_LIMIT
+            });
         let ctx = SessionContext::new_with_config_rt(
             SessionConfig::default(),
             RuntimeEnvBuilder::new()
-                .with_memory_limit(100 * 1024 * 1024, 1.0)
+                .with_memory_limit(memory_limit, 1.0)
                 .with_disk_manager_builder(
                     DiskManagerBuilder::default()
                         .with_mode(self.config.temp_dir.to_disk_manager_mode()),
@@ -232,7 +244,7 @@ impl PermutationBuilder {
     /// Builds the permutation table and stores it in the given database.
     pub async fn build(self) -> Result<Table> {
         // First pass, apply filter and load row ids
-        let mut rows = self.base_table.query().with_row_id();
+        let mut rows = self.base_table.query().select(Select::columns(&[ROW_ID]));
 
         if let Some(filter) = &self.config.filter {
             rows = rows.only_if(filter);
@@ -321,6 +333,47 @@ mod tests {
 
     use super::*;
 
+    #[tokio::test]
+    async fn test_permutation_table_only_stores_row_id_and_split_id() {
+        let temp_dir = tempfile::tempdir().unwrap();
+
+        let db = connect(temp_dir.path().to_str().unwrap())
+            .execute()
+            .await
+            .unwrap();
+
+        let initial_data = lance_datagen::gen_batch()
+            .col("col_a", lance_datagen::array::step::<Int32Type>())
+            .col("col_b", lance_datagen::array::step::<Int32Type>())
+            .into_ldb_stream(RowCount::from(100), BatchCount::from(10));
+        let data_table = db
+            .create_table_streaming("base_tbl", initial_data)
+            .execute()
+            .await
+            .unwrap();
+
+        let permutation_table = PermutationBuilder::new(data_table.clone())
+            .with_split_strategy(
+                SplitStrategy::Sequential {
+                    sizes: SplitSizes::Percentages(vec![0.5, 0.5]),
+                },
+                None,
+            )
+            .with_filter("col_a > 57".to_string())
+            .build()
+            .await
+            .unwrap();
+
+        let schema = permutation_table.schema().await.unwrap();
+        let field_names: Vec<&str> = schema.fields().iter().map(|f| f.name().as_str()).collect();
+        assert_eq!(
+            field_names,
+            vec!["row_id", "split_id"],
+            "Permutation table should only contain row_id and split_id columns, but found: {:?}",
+            field_names,
+        );
+    }
+
     #[tokio::test]
     async fn test_permutation_builder() {
         let temp_dir = tempfile::tempdir().unwrap();
@@ -352,8 +405,6 @@ mod tests {
             .await
             .unwrap();
 
-        println!("permutation_table: {:?}", permutation_table);
-
         // Potentially brittle seed-dependent values below
         assert_eq!(permutation_table.count_rows(None).await.unwrap(), 330);
         assert_eq!(

rust/lancedb/src/dataloader/permutation/shuffle.rs

@@ -171,7 +171,7 @@ impl Shuffler {
             // This is kind of an annoying limitation but if we allow runt clumps from batches then
             // clumps will get unaligned and we will mess up the clumps when we do the in-memory
             // shuffle step.  If this is a problem we can probably figure out a better way to do this.
-            if !is_last && batch.num_rows() as u64 % clump_size != 0 {
+            if !is_last && !(batch.num_rows() as u64).is_multiple_of(clump_size) {
                 return Err(Error::Runtime {
                     message: format!(
                         "Expected batch size ({}) to be divisible by clump size ({})",

rust/lancedb/src/dataloader/permutation/split.rs

@@ -1,12 +1,9 @@
 // SPDX-License-Identifier: Apache-2.0
 // SPDX-FileCopyrightText: Copyright The LanceDB Authors
 
-use std::{
-    iter,
-    sync::{
-        atomic::{AtomicBool, AtomicU64, AtomicUsize, Ordering},
-        Arc,
-    },
+use std::sync::{
+    atomic::{AtomicBool, AtomicU64, AtomicUsize, Ordering},
+    Arc,
 };
 
 use arrow_array::{Array, BooleanArray, RecordBatch, UInt64Array};
@@ -15,6 +12,8 @@ use datafusion_common::hash_utils::create_hashes;
 use futures::{StreamExt, TryStreamExt};
 use lance_arrow::SchemaExt;
 
+use lance_core::ROW_ID;
+
 use crate::{
     arrow::{SendableRecordBatchStream, SimpleRecordBatchStream},
     dataloader::{
@@ -158,7 +157,7 @@ impl Splitter {
                 remaining_in_split
             };
 
-            split_ids.extend(iter::repeat(split_id as u64).take(rows_to_add as usize));
+            split_ids.extend(std::iter::repeat_n(split_id as u64, rows_to_add as usize));
             if done {
                 // Quit early if we've run out of splits
                 break;
@@ -363,11 +362,15 @@ impl Splitter {
 
     pub fn project(&self, query: Query) -> Query {
         match &self.strategy {
-            SplitStrategy::Calculated { calculation } => query.select(Select::Dynamic(vec![(
-                SPLIT_ID_COLUMN.to_string(),
-                calculation.clone(),
-            )])),
-            SplitStrategy::Hash { columns, .. } => query.select(Select::Columns(columns.clone())),
+            SplitStrategy::Calculated { calculation } => query.select(Select::Dynamic(vec![
+                (SPLIT_ID_COLUMN.to_string(), calculation.clone()),
+                (ROW_ID.to_string(), ROW_ID.to_string()),
+            ])),
+            SplitStrategy::Hash { columns, .. } => {
+                let mut cols = columns.clone();
+                cols.push(ROW_ID.to_string());
+                query.select(Select::Columns(cols))
+            }
             _ => query,
         }
     }
@@ -662,7 +665,7 @@ mod tests {
         assert_eq!(split_batch.num_rows(), total_split_sizes as usize);
         let mut expected = Vec::with_capacity(total_split_sizes as usize);
         for (i, size) in expected_split_sizes.iter().enumerate() {
-            expected.extend(iter::repeat(i as u64).take(*size as usize));
+            expected.extend(std::iter::repeat_n(i as u64, *size as usize));
         }
         let expected = Arc::new(UInt64Array::from(expected)) as Arc<dyn Array>;
 

rust/lancedb/src/embeddings.rs

@@ -120,8 +120,13 @@ impl MemoryRegistry {
 }
 
 /// A record batch reader that has embeddings applied to it
-/// This is a wrapper around another record batch reader that applies an embedding function
-/// when reading from the record batch
+///
+/// This is a wrapper around another record batch reader that applies embedding functions
+/// when reading from the record batch.
+///
+/// When multiple embedding functions are defined, they are computed in parallel using
+/// scoped threads to improve performance. For a single embedding function, computation
+/// is done inline without threading overhead.
 pub struct WithEmbeddings<R: RecordBatchReader> {
     inner: R,
     embeddings: Vec<(EmbeddingDefinition, Arc<dyn EmbeddingFunction>)>,
@@ -235,6 +240,48 @@ impl<R: RecordBatchReader> WithEmbeddings<R> {
             column_definitions,
         })
     }
+
+    fn compute_embeddings_parallel(&self, batch: &RecordBatch) -> Result<Vec<Arc<dyn Array>>> {
+        if self.embeddings.len() == 1 {
+            let (fld, func) = &self.embeddings[0];
+            let src_column =
+                batch
+                    .column_by_name(&fld.source_column)
+                    .ok_or_else(|| Error::InvalidInput {
+                        message: format!("Source column '{}' not found", fld.source_column),
+                    })?;
+            return Ok(vec![func.compute_source_embeddings(src_column.clone())?]);
+        }
+
+        // Parallel path: multiple embeddings
+        std::thread::scope(|s| {
+            let handles: Vec<_> = self
+                .embeddings
+                .iter()
+                .map(|(fld, func)| {
+                    let src_column = batch.column_by_name(&fld.source_column).ok_or_else(|| {
+                        Error::InvalidInput {
+                            message: format!("Source column '{}' not found", fld.source_column),
+                        }
+                    })?;
+
+                    let handle =
+                        s.spawn(move || func.compute_source_embeddings(src_column.clone()));
+
+                    Ok(handle)
+                })
+                .collect::<Result<_>>()?;
+
+            handles
+                .into_iter()
+                .map(|h| {
+                    h.join().map_err(|e| Error::Runtime {
+                        message: format!("Thread panicked during embedding computation: {:?}", e),
+                    })?
+                })
+                .collect()
+        })
+    }
 }
 
 impl<R: RecordBatchReader> Iterator for MaybeEmbedded<R> {
@@ -262,19 +309,19 @@ impl<R: RecordBatchReader> Iterator for WithEmbeddings<R> {
     fn next(&mut self) -> Option<Self::Item> {
         let batch = self.inner.next()?;
         match batch {
-            Ok(mut batch) => {
-                // todo: parallelize this
-                for (fld, func) in self.embeddings.iter() {
-                    let src_column = batch.column_by_name(&fld.source_column).unwrap();
-                    let embedding = match func.compute_source_embeddings(src_column.clone()) {
-                        Ok(embedding) => embedding,
-                        Err(e) => {
-                            return Some(Err(arrow_schema::ArrowError::ComputeError(format!(
-                                "Error computing embedding: {}",
-                                e
-                            ))))
-                        }
-                    };
+            Ok(batch) => {
+                let embeddings = match self.compute_embeddings_parallel(&batch) {
+                    Ok(emb) => emb,
+                    Err(e) => {
+                        return Some(Err(arrow_schema::ArrowError::ComputeError(format!(
+                            "Error computing embedding: {}",
+                            e
+                        ))))
+                    }
+                };
+
+                let mut batch = batch;
+                for ((fld, _), embedding) in self.embeddings.iter().zip(embeddings.iter()) {
                     let dst_field_name = fld
                         .dest_column
                         .clone()
@@ -286,7 +333,7 @@ impl<R: RecordBatchReader> Iterator for WithEmbeddings<R> {
                         embedding.nulls().is_some(),
                     );
 
-                    match batch.try_with_column(dst_field.clone(), embedding) {
+                    match batch.try_with_column(dst_field.clone(), embedding.clone()) {
                         Ok(b) => batch = b,
                         Err(e) => return Some(Err(e)),
                     };

rust/lancedb/src/index/vector.rs

@@ -297,10 +297,10 @@ impl IvfPqIndexBuilder {
 }
 
 pub(crate) fn suggested_num_sub_vectors(dim: u32) -> u32 {
-    if dim % 16 == 0 {
+    if dim.is_multiple_of(16) {
         // Should be more aggressive than this default.
         dim / 16
-    } else if dim % 8 == 0 {
+    } else if dim.is_multiple_of(8) {
         dim / 8
     } else {
         log::warn!(

rust/lancedb/src/lib.rs

@@ -25,13 +25,14 @@
 //!
 //! ## Crate Features
 //!
-//! ### Experimental Features
-//!
-//! These features are not enabled by default.  They are experimental or in-development features that
-//! are not yet ready to be released.
-//!
-//! - `remote` - Enable remote client to connect to LanceDB cloud.  This is not yet fully implemented
-//!   and should not be enabled.
+//! - `aws` - Enable AWS S3 object store support.
+//! - `dynamodb` - Enable DynamoDB manifest store support.
+//! - `azure` - Enable Azure Blob Storage object store support.
+//! - `gcs` - Enable Google Cloud Storage object store support.
+//! - `oss` - Enable Alibaba Cloud OSS object store support.
+//! - `remote` - Enable remote client to connect to LanceDB cloud.
+//! - `huggingface` - Enable HuggingFace Hub integration for loading datasets from the Hub.
+//! - `fp16kernels` - Enable FP16 kernels for faster vector search on CPU.
 //!
 //! ### Quick Start
 //!
@@ -50,17 +51,15 @@
 //! - `s3://bucket/path/to/database` or `gs://bucket/path/to/database` - database on cloud object store
 //! - `db://dbname` - Lance Cloud
 //!
-//! You can also use [`ConnectOptions`] to configure the connection to the database.
+//! You can also use [`ConnectBuilder`] to configure the connection to the database.
 //!
 //! ```rust
-//! use object_store::aws::AwsCredential;
 //! # tokio::runtime::Runtime::new().unwrap().block_on(async {
 //! let db = lancedb::connect("data/sample-lancedb")
-//!     .aws_creds(AwsCredential {
-//!         key_id: "some_key".to_string(),
-//!         secret_key: "some_secret".to_string(),
-//!         token: None,
-//!     })
+//!     .storage_options([
+//!         ("aws_access_key_id", "some_key"),
+//!         ("aws_secret_access_key", "some_secret"),
+//!     ])
 //!     .execute()
 //!     .await
 //!     .unwrap();

rust/lancedb/src/remote/client.rs

@@ -232,6 +232,38 @@ impl HttpSend for Sender {
     }
 }
 
+/// Parsed components from a database URL (db://...)
+pub struct ParsedDbUrl {
+    pub db_name: String,
+    pub db_prefix: Option<String>,
+}
+
+/// Parse a database URL and extract the database name and optional prefix.
+///
+/// Expected format: `db://db_name` or `db://db_name/prefix`
+pub fn parse_db_url(db_url: &str) -> Result<ParsedDbUrl> {
+    let parsed_url = url::Url::parse(db_url).map_err(|err| Error::InvalidInput {
+        message: format!("db_url is not a valid URL. '{db_url}'. Error: {err}"),
+    })?;
+    debug_assert_eq!(parsed_url.scheme(), "db");
+    if !parsed_url.has_host() {
+        return Err(Error::InvalidInput {
+            message: format!("Invalid database URL (missing host) '{}'", db_url),
+        });
+    }
+    let db_name = parsed_url.host_str().unwrap().to_string();
+    let db_prefix = {
+        let prefix = parsed_url.path().trim_start_matches('/');
+        if prefix.is_empty() {
+            None
+        } else {
+            Some(prefix.to_string())
+        }
+    };
+
+    Ok(ParsedDbUrl { db_name, db_prefix })
+}
+
 impl RestfulLanceDbClient<Sender> {
     fn get_timeout(passed: Option<Duration>, env_var: &str) -> Result<Option<Duration>> {
         if let Some(passed) = passed {
@@ -250,32 +282,12 @@ impl RestfulLanceDbClient<Sender> {
     }
 
     pub fn try_new(
-        db_url: &str,
-        api_key: &str,
+        parsed_url: &ParsedDbUrl,
         region: &str,
         host_override: Option<String>,
+        default_headers: HeaderMap,
         client_config: ClientConfig,
-        options: &RemoteOptions,
     ) -> Result<Self> {
-        let parsed_url = url::Url::parse(db_url).map_err(|err| Error::InvalidInput {
-            message: format!("db_url is not a valid URL. '{db_url}'. Error: {err}"),
-        })?;
-        debug_assert_eq!(parsed_url.scheme(), "db");
-        if !parsed_url.has_host() {
-            return Err(Error::InvalidInput {
-                message: format!("Invalid database URL (missing host) '{}'", db_url),
-            });
-        }
-        let db_name = parsed_url.host_str().unwrap();
-        let db_prefix = {
-            let prefix = parsed_url.path().trim_start_matches('/');
-            if prefix.is_empty() {
-                None
-            } else {
-                Some(prefix)
-            }
-        };
-
         // Get the timeouts
         let timeout =
             Self::get_timeout(client_config.timeout_config.timeout, "LANCE_CLIENT_TIMEOUT")?;
@@ -348,15 +360,7 @@ impl RestfulLanceDbClient<Sender> {
         }
 
         let client = client_builder
-            .default_headers(Self::default_headers(
-                api_key,
-                region,
-                db_name,
-                host_override.is_some(),
-                options,
-                db_prefix,
-                &client_config,
-            )?)
+            .default_headers(default_headers)
             .user_agent(client_config.user_agent)
             .build()
             .map_err(|err| Error::Other {
@@ -366,7 +370,7 @@ impl RestfulLanceDbClient<Sender> {
 
         let host = match host_override {
             Some(host_override) => host_override,
-            None => format!("https://{}.{}.api.lancedb.com", db_name, region),
+            None => format!("https://{}.{}.api.lancedb.com", parsed_url.db_name, region),
         };
         debug!("Created client for host: {}", host);
         let retry_config = client_config.retry_config.clone().try_into()?;
@@ -389,7 +393,7 @@ impl<S: HttpSend> RestfulLanceDbClient<S> {
         &self.host
     }
 
-    fn default_headers(
+    pub fn default_headers(
         api_key: &str,
         region: &str,
         db_name: &str,

rust/lancedb/src/remote/db.rs

@@ -189,6 +189,10 @@ pub struct RemoteDatabase<S: HttpSend = Sender> {
     client: RestfulLanceDbClient<S>,
     table_cache: Cache<String, Arc<RemoteTable<S>>>,
     uri: String,
+    /// Headers to pass to the namespace client for authentication
+    namespace_headers: HashMap<String, String>,
+    /// TLS configuration for mTLS support
+    tls_config: Option<super::client::TlsConfig>,
 }
 
 impl RemoteDatabase {
@@ -200,13 +204,32 @@ impl RemoteDatabase {
         client_config: ClientConfig,
         options: RemoteOptions,
     ) -> Result<Self> {
-        let client = RestfulLanceDbClient::try_new(
-            uri,
+        let parsed = super::client::parse_db_url(uri)?;
+        let header_map = RestfulLanceDbClient::<Sender>::default_headers(
             api_key,
             region,
-            host_override,
-            client_config,
+            &parsed.db_name,
+            host_override.is_some(),
             &options,
+            parsed.db_prefix.as_deref(),
+            &client_config,
+        )?;
+
+        let namespace_headers: HashMap<String, String> = header_map
+            .iter()
+            .filter_map(|(k, v)| {
+                v.to_str()
+                    .ok()
+                    .map(|val| (k.as_str().to_string(), val.to_string()))
+            })
+            .collect();
+
+        let client = RestfulLanceDbClient::try_new(
+            &parsed,
+            region,
+            host_override,
+            header_map,
+            client_config.clone(),
         )?;
 
         let table_cache = Cache::builder()
@@ -218,6 +241,8 @@ impl RemoteDatabase {
             client,
             table_cache,
             uri: uri.to_owned(),
+            namespace_headers,
+            tls_config: client_config.tls_config,
         })
     }
 }
@@ -240,6 +265,8 @@ mod test_utils {
                 client,
                 table_cache: Cache::new(0),
                 uri: "http://localhost".to_string(),
+                namespace_headers: HashMap::new(),
+                tls_config: None,
             }
         }
 
@@ -248,11 +275,13 @@ mod test_utils {
             F: Fn(reqwest::Request) -> http::Response<T> + Send + Sync + 'static,
             T: Into<reqwest::Body>,
         {
-            let client = client_with_handler_and_config(handler, config);
+            let client = client_with_handler_and_config(handler, config.clone());
             Self {
                 client,
                 table_cache: Cache::new(0),
                 uri: "http://localhost".to_string(),
+                namespace_headers: config.extra_headers.clone(),
+                tls_config: config.tls_config.clone(),
             }
         }
     }
@@ -716,7 +745,8 @@ impl<S: HttpSend> Database for RemoteDatabase<S> {
         let namespace_id = build_namespace_identifier(namespace_parts, &self.client.id_delimiter);
         let req = self
             .client
-            .get(&format!("/v1/namespace/{}/describe", namespace_id));
+            .post(&format!("/v1/namespace/{}/describe", namespace_id))
+            .json(&DescribeNamespaceRequest::default());
 
         let (request_id, resp) = self.client.send(req).await?;
         let resp = self.client.check_response(&request_id, resp).await?;
@@ -727,6 +757,31 @@ impl<S: HttpSend> Database for RemoteDatabase<S> {
     fn as_any(&self) -> &dyn std::any::Any {
         self
     }
+
+    async fn namespace_client(&self) -> Result<Arc<dyn lance_namespace::LanceNamespace>> {
+        // Create a RestNamespace pointing to the same remote host with the same authentication headers
+        let mut builder = lance_namespace_impls::RestNamespaceBuilder::new(self.client.host())
+            .delimiter(&self.client.id_delimiter)
+            // TODO: support header provider
+            .headers(self.namespace_headers.clone());
+
+        // Apply mTLS configuration if present
+        if let Some(tls_config) = &self.tls_config {
+            if let Some(cert_file) = &tls_config.cert_file {
+                builder = builder.cert_file(cert_file);
+            }
+            if let Some(key_file) = &tls_config.key_file {
+                builder = builder.key_file(key_file);
+            }
+            if let Some(ssl_ca_cert) = &tls_config.ssl_ca_cert {
+                builder = builder.ssl_ca_cert(ssl_ca_cert);
+            }
+            builder = builder.assert_hostname(tls_config.assert_hostname);
+        }
+
+        let namespace = builder.build();
+        Ok(Arc::new(namespace) as Arc<dyn lance_namespace::LanceNamespace>)
+    }
 }
 
 /// RemoteOptions contains a subset of StorageOptions that are compatible with Remote LanceDB connections
@@ -1518,4 +1573,260 @@ mod tests {
             panic!("Expected HTTP error");
         }
     }
+
+    #[tokio::test]
+    async fn test_namespace_client() {
+        let conn = Connection::new_with_handler(|_| {
+            http::Response::builder()
+                .status(200)
+                .body(r#"{"tables": []}"#)
+                .unwrap()
+        });
+
+        // Get the namespace client from the connection's internal database
+        let namespace_client = conn.namespace_client().await;
+        assert!(namespace_client.is_ok());
+    }
+
+    #[tokio::test]
+    async fn test_namespace_client_with_tls_config() {
+        use crate::remote::client::TlsConfig;
+
+        let tls_config = TlsConfig {
+            cert_file: Some("/path/to/cert.pem".to_string()),
+            key_file: Some("/path/to/key.pem".to_string()),
+            ssl_ca_cert: Some("/path/to/ca.pem".to_string()),
+            assert_hostname: true,
+        };
+
+        let client_config = ClientConfig {
+            tls_config: Some(tls_config),
+            ..Default::default()
+        };
+
+        let conn = Connection::new_with_handler_and_config(
+            |_| {
+                http::Response::builder()
+                    .status(200)
+                    .body(r#"{"tables": []}"#)
+                    .unwrap()
+            },
+            client_config,
+        );
+
+        // Get the namespace client - it should be created with the TLS config
+        let namespace_client = conn.namespace_client().await;
+        assert!(namespace_client.is_ok());
+    }
+
+    #[tokio::test]
+    async fn test_namespace_client_with_headers() {
+        let mut extra_headers = HashMap::new();
+        extra_headers.insert("X-Custom-Header".to_string(), "custom-value".to_string());
+
+        let client_config = ClientConfig {
+            extra_headers,
+            ..Default::default()
+        };
+
+        let conn = Connection::new_with_handler_and_config(
+            |_| {
+                http::Response::builder()
+                    .status(200)
+                    .body(r#"{"tables": []}"#)
+                    .unwrap()
+            },
+            client_config,
+        );
+
+        // Get the namespace client - it should be created with the extra headers
+        let namespace_client = conn.namespace_client().await;
+        assert!(namespace_client.is_ok());
+    }
+
+    /// Integration tests using RestAdapter to run RemoteDatabase against a real namespace server
+    mod rest_adapter_integration {
+        use super::*;
+        use lance_namespace::models::ListTablesRequest;
+        use lance_namespace_impls::{DirectoryNamespaceBuilder, RestAdapter, RestAdapterConfig};
+        use std::sync::Arc;
+        use tempfile::TempDir;
+
+        /// Test fixture that manages a REST server backed by DirectoryNamespace
+        struct RestServerFixture {
+            _temp_dir: TempDir,
+            server_handle: lance_namespace_impls::RestAdapterHandle,
+            server_url: String,
+        }
+
+        impl RestServerFixture {
+            async fn new() -> Self {
+                let temp_dir = TempDir::new().unwrap();
+                let temp_path = temp_dir.path().to_str().unwrap().to_string();
+
+                // Create DirectoryNamespace backend
+                let backend = DirectoryNamespaceBuilder::new(&temp_path)
+                    .build()
+                    .await
+                    .unwrap();
+                let backend = Arc::new(backend);
+
+                // Start REST server with port 0 (OS assigns available port)
+                let config = RestAdapterConfig {
+                    port: 0,
+                    ..Default::default()
+                };
+
+                let server = RestAdapter::new(backend, config);
+                let server_handle = server.start().await.unwrap();
+
+                // Get the actual port assigned by OS
+                let actual_port = server_handle.port();
+                let server_url = format!("http://127.0.0.1:{}", actual_port);
+
+                Self {
+                    _temp_dir: temp_dir,
+                    server_handle,
+                    server_url,
+                }
+            }
+        }
+
+        impl Drop for RestServerFixture {
+            fn drop(&mut self) {
+                self.server_handle.shutdown();
+            }
+        }
+
+        #[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+        async fn test_remote_database_with_rest_adapter() {
+            use lance_namespace::models::CreateNamespaceRequest;
+
+            let fixture = RestServerFixture::new().await;
+
+            // Connect to the REST server using lancedb Connection
+            // Use db://dummy as URI and set actual server URL via host_override
+            let conn = ConnectBuilder::new("db://dummy")
+                .api_key("test-api-key")
+                .region("us-east-1")
+                .host_override(&fixture.server_url)
+                .execute()
+                .await
+                .unwrap();
+
+            // Create a child namespace first
+            let namespace = vec!["test_ns".to_string()];
+            conn.create_namespace(CreateNamespaceRequest {
+                id: Some(namespace.clone()),
+                ..Default::default()
+            })
+            .await
+            .expect("Failed to create namespace");
+
+            // Create a table in the child namespace
+            let schema = Arc::new(Schema::new(vec![Field::new("a", DataType::Int32, false)]));
+            let data = RecordBatch::try_new(
+                schema.clone(),
+                vec![Arc::new(Int32Array::from(vec![1, 2, 3]))],
+            )
+            .unwrap();
+            let reader = RecordBatchIterator::new([Ok(data.clone())], schema.clone());
+
+            let table = conn
+                .create_table("test_table", reader)
+                .namespace(namespace.clone())
+                .execute()
+                .await;
+            assert!(table.is_ok(), "Failed to create table: {:?}", table.err());
+
+            // List tables in the child namespace
+            let list_response = conn
+                .list_tables(ListTablesRequest {
+                    id: Some(namespace.clone()),
+                    ..Default::default()
+                })
+                .await
+                .expect("Failed to list tables");
+            assert_eq!(list_response.tables, vec!["test_table"]);
+
+            // Get namespace client and verify it can also list tables
+            let namespace_client = conn.namespace_client().await.unwrap();
+            let list_response = namespace_client
+                .list_tables(ListTablesRequest {
+                    id: Some(namespace.clone()),
+                    ..Default::default()
+                })
+                .await
+                .unwrap();
+            assert_eq!(list_response.tables, vec!["test_table"]);
+
+            // Open the table from the child namespace
+            let opened_table = conn
+                .open_table("test_table")
+                .namespace(namespace.clone())
+                .execute()
+                .await;
+            assert!(
+                opened_table.is_ok(),
+                "Failed to open table: {:?}",
+                opened_table.err()
+            );
+            assert_eq!(opened_table.unwrap().name(), "test_table");
+        }
+
+        #[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+        async fn test_remote_database_with_multiple_tables() {
+            use lance_namespace::models::CreateNamespaceRequest;
+
+            let fixture = RestServerFixture::new().await;
+
+            // Connect to the REST server
+            // Use db://dummy as URI and set actual server URL via host_override
+            let conn = ConnectBuilder::new("db://dummy")
+                .api_key("test-api-key")
+                .region("us-east-1")
+                .host_override(&fixture.server_url)
+                .execute()
+                .await
+                .unwrap();
+
+            // Create a child namespace first
+            let namespace = vec!["multi_table_ns".to_string()];
+            conn.create_namespace(CreateNamespaceRequest {
+                id: Some(namespace.clone()),
+                ..Default::default()
+            })
+            .await
+            .expect("Failed to create namespace");
+
+            // Create multiple tables in the child namespace
+            let schema = Arc::new(Schema::new(vec![Field::new("id", DataType::Int32, false)]));
+
+            for i in 1..=3 {
+                let data =
+                    RecordBatch::try_new(schema.clone(), vec![Arc::new(Int32Array::from(vec![i]))])
+                        .unwrap();
+                let reader = RecordBatchIterator::new([Ok(data.clone())], schema.clone());
+
+                conn.create_table(format!("table{}", i), reader)
+                    .namespace(namespace.clone())
+                    .execute()
+                    .await
+                    .unwrap_or_else(|e| panic!("Failed to create table{}: {:?}", i, e));
+            }
+
+            // List tables in the child namespace
+            let list_response = conn
+                .list_tables(ListTablesRequest {
+                    id: Some(namespace.clone()),
+                    ..Default::default()
+                })
+                .await
+                .unwrap();
+            assert_eq!(list_response.tables.len(), 3);
+            assert!(list_response.tables.contains(&"table1".to_string()));
+            assert!(list_response.tables.contains(&"table2".to_string()));
+            assert!(list_response.tables.contains(&"table3".to_string()));
+        }
+    }
 }

rust/lancedb/src/remote/table.rs

@@ -204,6 +204,7 @@ pub struct RemoteTable<S: HttpSend = Sender> {
     server_version: ServerVersion,
 
     version: RwLock<Option<u64>>,
+    location: RwLock<Option<String>>,
 }
 
 impl<S: HttpSend> RemoteTable<S> {
@@ -221,6 +222,7 @@ impl<S: HttpSend> RemoteTable<S> {
             identifier,
             server_version,
             version: RwLock::new(None),
+            location: RwLock::new(None),
         }
     }
 
@@ -639,6 +641,7 @@ impl<S: HttpSend> RemoteTable<S> {
 struct TableDescription {
     version: u64,
     schema: JsonSchema,
+    location: Option<String>,
 }
 
 impl<S: HttpSend> std::fmt::Display for RemoteTable<S> {
@@ -667,6 +670,7 @@ mod test_utils {
                 identifier: name,
                 server_version: version.map(ServerVersion).unwrap_or_default(),
                 version: RwLock::new(None),
+                location: RwLock::new(None),
             }
         }
     }
@@ -1088,6 +1092,17 @@ impl<S: HttpSend> BaseTable for RemoteTable<S> {
                     body["num_partitions"] = serde_json::Value::Number(num_partitions.into());
                 }
             }
+            Index::IvfRq(index) => {
+                body[INDEX_TYPE_KEY] = serde_json::Value::String("IVF_RQ".to_string());
+                body[METRIC_TYPE_KEY] =
+                    serde_json::Value::String(index.distance_type.to_string().to_lowercase());
+                if let Some(num_partitions) = index.num_partitions {
+                    body["num_partitions"] = serde_json::Value::Number(num_partitions.into());
+                }
+                if let Some(num_bits) = index.num_bits {
+                    body["num_bits"] = serde_json::Value::Number(num_bits.into());
+                }
+            }
             Index::BTree(_) => {
                 body[INDEX_TYPE_KEY] = serde_json::Value::String("BTREE".to_string());
             }
@@ -1450,8 +1465,28 @@ impl<S: HttpSend> BaseTable for RemoteTable<S> {
             message: "table_definition is not supported on LanceDB cloud.".into(),
         })
     }
-    fn dataset_uri(&self) -> &str {
-        "NOT_SUPPORTED"
+    async fn uri(&self) -> Result<String> {
+        // Check if we already have the location cached
+        {
+            let location = self.location.read().await;
+            if let Some(ref loc) = *location {
+                return Ok(loc.clone());
+            }
+        }
+
+        // Fetch from server via describe
+        let description = self.describe().await?;
+        let location = description.location.ok_or_else(|| Error::NotSupported {
+            message: "Table URI not supported by the server".into(),
+        })?;
+
+        // Cache the location for future use
+        {
+            let mut cached_location = self.location.write().await;
+            *cached_location = Some(location.clone());
+        }
+
+        Ok(location)
     }
 
     async fn storage_options(&self) -> Option<HashMap<String, String>> {
@@ -3321,4 +3356,69 @@ mod tests {
         let result = table.drop_columns(&["old_col1", "old_col2"]).await.unwrap();
         assert_eq!(result.version, 5);
     }
+
+    #[tokio::test]
+    async fn test_uri() {
+        let table = Table::new_with_handler("my_table", |request| {
+            assert_eq!(request.method(), "POST");
+            assert_eq!(request.url().path(), "/v1/table/my_table/describe/");
+
+            http::Response::builder()
+                .status(200)
+                .body(r#"{"version": 1, "schema": {"fields": []}, "location": "s3://bucket/path/to/table"}"#)
+                .unwrap()
+        });
+
+        let uri = table.uri().await.unwrap();
+        assert_eq!(uri, "s3://bucket/path/to/table");
+    }
+
+    #[tokio::test]
+    async fn test_uri_missing_location() {
+        let table = Table::new_with_handler("my_table", |request| {
+            assert_eq!(request.method(), "POST");
+            assert_eq!(request.url().path(), "/v1/table/my_table/describe/");
+
+            // Server returns response without location field
+            http::Response::builder()
+                .status(200)
+                .body(r#"{"version": 1, "schema": {"fields": []}}"#)
+                .unwrap()
+        });
+
+        let result = table.uri().await;
+        assert!(result.is_err());
+        assert!(matches!(&result, Err(Error::NotSupported { .. })));
+    }
+
+    #[tokio::test]
+    async fn test_uri_caching() {
+        use std::sync::atomic::{AtomicUsize, Ordering};
+        use std::sync::Arc;
+
+        let call_count = Arc::new(AtomicUsize::new(0));
+        let call_count_clone = call_count.clone();
+
+        let table = Table::new_with_handler("my_table", move |request| {
+            assert_eq!(request.url().path(), "/v1/table/my_table/describe/");
+            call_count_clone.fetch_add(1, Ordering::SeqCst);
+
+            http::Response::builder()
+                .status(200)
+                .body(
+                    r#"{"version": 1, "schema": {"fields": []}, "location": "gs://bucket/table"}"#,
+                )
+                .unwrap()
+        });
+
+        // First call should fetch from server
+        let uri1 = table.uri().await.unwrap();
+        assert_eq!(uri1, "gs://bucket/table");
+        assert_eq!(call_count.load(Ordering::SeqCst), 1);
+
+        // Second call should use cached value
+        let uri2 = table.uri().await.unwrap();
+        assert_eq!(uri2, "gs://bucket/table");
+        assert_eq!(call_count.load(Ordering::SeqCst), 1); // Still 1, no new call
+    }
 }

rust/lancedb/src/table.rs

@@ -29,7 +29,7 @@ use lance::dataset::{
 use lance::dataset::{MergeInsertBuilder as LanceMergeInsertBuilder, WhenNotMatchedBySource};
 use lance::index::vector::utils::infer_vector_dim;
 use lance::index::vector::VectorIndexParams;
-use lance::io::WrappingObjectStore;
+use lance::io::{ObjectStoreParams, WrappingObjectStore};
 use lance_datafusion::exec::{analyze_plan as lance_analyze_plan, execute_plan};
 use lance_datafusion::utils::StreamingWriteSource;
 use lance_index::scalar::{BuiltinIndexType, ScalarIndexParams};
@@ -40,9 +40,10 @@ use lance_index::vector::pq::PQBuildParams;
 use lance_index::vector::sq::builder::SQBuildParams;
 use lance_index::DatasetIndexExt;
 use lance_index::IndexType;
+use lance_io::object_store::{LanceNamespaceStorageOptionsProvider, StorageOptionsAccessor};
 use lance_namespace::models::{
-    QueryTableRequest as NsQueryTableRequest, QueryTableRequestFullTextQuery,
-    QueryTableRequestVector, StringFtsQuery,
+    QueryTableRequest as NsQueryTableRequest, QueryTableRequestColumns,
+    QueryTableRequestFullTextQuery, QueryTableRequestVector, StringFtsQuery,
 };
 use lance_namespace::LanceNamespace;
 use lance_table::format::Manifest;
@@ -78,10 +79,11 @@ use self::merge::MergeInsertBuilder;
 
 pub mod datafusion;
 pub(crate) mod dataset;
+pub mod delete;
 pub mod merge;
-
 use crate::index::waiter::wait_for_index;
 pub use chrono::Duration;
+pub use delete::DeleteResult;
 use futures::future::{join_all, Either};
 pub use lance::dataset::optimize::CompactionOptions;
 pub use lance::dataset::refs::{TagContents, Tags as LanceTags};
@@ -445,15 +447,6 @@ pub struct AddResult {
     pub version: u64,
 }
 
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize, Default)]
-pub struct DeleteResult {
-    // The commit version associated with the operation.
-    // A version of `0` indicates compatibility with legacy servers that do not return
-    /// a commit version.
-    #[serde(default)]
-    pub version: u64,
-}
-
 #[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize, Default)]
 pub struct MergeResult {
     // The commit version associated with the operation.
@@ -607,8 +600,8 @@ pub trait BaseTable: std::fmt::Display + std::fmt::Debug + Send + Sync {
     async fn list_versions(&self) -> Result<Vec<Version>>;
     /// Get the table definition.
     async fn table_definition(&self) -> Result<TableDefinition>;
-    /// Get the table URI
-    fn dataset_uri(&self) -> &str;
+    /// Get the table URI (storage location)
+    async fn uri(&self) -> Result<String>;
     /// Get the storage options used when opening this table, if any.
     async fn storage_options(&self) -> Option<HashMap<String, String>>;
     /// Poll until the columns are fully indexed. Will return Error::Timeout if the columns
@@ -1316,11 +1309,12 @@ impl Table {
         self.inner.list_indices().await
     }
 
-    /// Get the underlying dataset URI
+    /// Get the table URI (storage location)
     ///
-    /// Warning: This is an internal API and the return value is subject to change.
-    pub fn dataset_uri(&self) -> &str {
-        self.inner.dataset_uri()
+    /// Returns the full storage location of the table (e.g., S3/GCS path).
+    /// For remote tables, this fetches the location from the server via describe.
+    pub async fn uri(&self) -> Result<String> {
+        self.inner.uri().await
     }
 
     /// Get the storage options used when opening this table, if any.
@@ -1611,6 +1605,101 @@ impl NativeTable {
         self
     }
 
+    /// Opens an existing Table using a namespace client.
+    ///
+    /// This method uses `DatasetBuilder::from_namespace` to open the table, which
+    /// automatically fetches the table location and storage options from the namespace.
+    /// This eliminates the need to pre-fetch and merge storage options before opening.
+    ///
+    /// # Arguments
+    ///
+    /// * `namespace_client` - The namespace client to use for fetching table metadata
+    /// * `name` - The table name
+    /// * `namespace` - The namespace path (e.g., vec!["parent", "child"])
+    /// * `write_store_wrapper` - Optional wrapper for the object store on write path
+    /// * `params` - Optional read parameters
+    /// * `read_consistency_interval` - Optional interval for read consistency
+    /// * `server_side_query_enabled` - Whether to enable server-side query execution.
+    ///   When true, the namespace_client will be stored and queries will be executed
+    ///   on the namespace server. When false, the namespace is only used for opening
+    ///   the table, and queries are executed locally.
+    /// * `session` - Optional session for object stores and caching
+    ///
+    /// # Returns
+    ///
+    /// * A [NativeTable] object.
+    #[allow(clippy::too_many_arguments)]
+    pub async fn open_from_namespace(
+        namespace_client: Arc<dyn LanceNamespace>,
+        name: &str,
+        namespace: Vec<String>,
+        write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
+        params: Option<ReadParams>,
+        read_consistency_interval: Option<std::time::Duration>,
+        server_side_query_enabled: bool,
+        session: Option<Arc<lance::session::Session>>,
+    ) -> Result<Self> {
+        let mut params = params.unwrap_or_default();
+
+        // Set the session in read params
+        if let Some(sess) = session {
+            params.session(sess);
+        }
+
+        // patch the params if we have a write store wrapper
+        let params = match write_store_wrapper.clone() {
+            Some(wrapper) => params.patch_with_store_wrapper(wrapper)?,
+            None => params,
+        };
+
+        // Build table_id from namespace + name
+        let mut table_id = namespace.clone();
+        table_id.push(name.to_string());
+
+        // Use DatasetBuilder::from_namespace which automatically fetches location
+        // and storage options from the namespace
+        let builder = DatasetBuilder::from_namespace(namespace_client.clone(), table_id)
+            .await
+            .map_err(|e| match e {
+                lance::Error::Namespace { source, .. } => Error::Runtime {
+                    message: format!("Failed to get table info from namespace: {:?}", source),
+                },
+                source => Error::Lance { source },
+            })?;
+
+        let dataset = builder
+            .with_read_params(params)
+            .load()
+            .await
+            .map_err(|e| match e {
+                lance::Error::DatasetNotFound { .. } => Error::TableNotFound {
+                    name: name.to_string(),
+                    source: Box::new(e),
+                },
+                source => Error::Lance { source },
+            })?;
+
+        let uri = dataset.uri().to_string();
+        let dataset = DatasetConsistencyWrapper::new_latest(dataset, read_consistency_interval);
+        let id = Self::build_id(&namespace, name);
+
+        let stored_namespace_client = if server_side_query_enabled {
+            Some(namespace_client)
+        } else {
+            None
+        };
+
+        Ok(Self {
+            name: name.to_string(),
+            namespace,
+            id,
+            uri,
+            dataset,
+            read_consistency_interval,
+            namespace_client: stored_namespace_client,
+        })
+    }
+
     fn get_table_name(uri: &str) -> Result<String> {
         let path = Path::new(uri);
         let name = path
@@ -1722,6 +1811,108 @@ impl NativeTable {
         .await
     }
 
+    /// Creates a new Table using a namespace client for storage options.
+    ///
+    /// This method sets up a `StorageOptionsProvider` from the namespace client,
+    /// enabling automatic credential refresh for cloud storage. The namespace
+    /// is used for:
+    /// 1. Setting up storage options provider for credential vending
+    /// 2. Optionally enabling server-side query execution
+    ///
+    /// # Arguments
+    ///
+    /// * `namespace_client` - The namespace client to use for storage options
+    /// * `uri` - The URI to the table (obtained from create_empty_table response)
+    /// * `name` - The table name
+    /// * `namespace` - The namespace path (e.g., vec!["parent", "child"])
+    /// * `batches` - RecordBatch to be saved in the database
+    /// * `write_store_wrapper` - Optional wrapper for the object store on write path
+    /// * `params` - Optional write parameters
+    /// * `read_consistency_interval` - Optional interval for read consistency
+    /// * `server_side_query_enabled` - Whether to enable server-side query execution
+    ///
+    /// # Returns
+    ///
+    /// * A [NativeTable] object.
+    #[allow(clippy::too_many_arguments)]
+    pub async fn create_from_namespace(
+        namespace_client: Arc<dyn LanceNamespace>,
+        uri: &str,
+        name: &str,
+        namespace: Vec<String>,
+        batches: impl StreamingWriteSource,
+        write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
+        params: Option<WriteParams>,
+        read_consistency_interval: Option<std::time::Duration>,
+        server_side_query_enabled: bool,
+        session: Option<Arc<lance::session::Session>>,
+    ) -> Result<Self> {
+        // Build table_id from namespace + name for the storage options provider
+        let mut table_id = namespace.clone();
+        table_id.push(name.to_string());
+
+        // Set up storage options provider from namespace
+        let storage_options_provider = Arc::new(LanceNamespaceStorageOptionsProvider::new(
+            namespace_client.clone(),
+            table_id,
+        ));
+
+        // Start with provided params or defaults
+        let mut params = params.unwrap_or_default();
+
+        // Set the session in write params
+        if let Some(sess) = session {
+            params.session = Some(sess);
+        }
+
+        // Ensure store_params exists and set the storage options provider
+        let store_params = params
+            .store_params
+            .get_or_insert_with(ObjectStoreParams::default);
+        let accessor = match store_params.storage_options().cloned() {
+            Some(options) => {
+                StorageOptionsAccessor::with_initial_and_provider(options, storage_options_provider)
+            }
+            None => StorageOptionsAccessor::with_provider(storage_options_provider),
+        };
+        store_params.storage_options_accessor = Some(Arc::new(accessor));
+
+        // Patch the params if we have a write store wrapper
+        let params = match write_store_wrapper.clone() {
+            Some(wrapper) => params.patch_with_store_wrapper(wrapper)?,
+            None => params,
+        };
+
+        let insert_builder = InsertBuilder::new(uri).with_params(&params);
+        let dataset = insert_builder
+            .execute_stream(batches)
+            .await
+            .map_err(|e| match e {
+                lance::Error::DatasetAlreadyExists { .. } => Error::TableAlreadyExists {
+                    name: name.to_string(),
+                },
+                source => Error::Lance { source },
+            })?;
+
+        let id = Self::build_id(&namespace, name);
+
+        let stored_namespace_client = if server_side_query_enabled {
+            Some(namespace_client)
+        } else {
+            None
+        };
+
+        Ok(Self {
+            name: name.to_string(),
+            namespace,
+            id,
+            uri: uri.to_string(),
+            dataset: DatasetConsistencyWrapper::new_latest(dataset, read_consistency_interval),
+            read_consistency_interval,
+            namespace_client: stored_namespace_client,
+        })
+    }
+
     async fn optimize_indices(&self, options: &OptimizeOptions) -> Result<()> {
         info!("LanceDB: optimizing indices: {:?}", options);
         self.dataset
@@ -1860,7 +2051,7 @@ impl NativeTable {
             return provided;
         }
         let suggested = suggested_num_sub_vectors(dim);
-        if num_bits.is_some_and(|num_bits| num_bits == 4) && suggested % 2 != 0 {
+        if num_bits.is_some_and(|num_bits| num_bits == 4) && !suggested.is_multiple_of(2) {
             // num_sub_vectors must be even when 4 bits are used
             suggested + 1
         } else {
@@ -2152,7 +2343,10 @@ impl NativeTable {
                 // Convert select to columns list
                 let columns = match &vq.base.select {
                     Select::All => None,
-                    Select::Columns(cols) => Some(cols.clone()),
+                    Select::Columns(cols) => Some(Box::new(QueryTableRequestColumns {
+                        column_names: Some(cols.clone()),
+                        column_aliases: None,
+                    })),
                     Select::Dynamic(_) => {
                         return Err(Error::NotSupported {
                             message:
@@ -2205,7 +2399,7 @@ impl NativeTable {
                     with_row_id: Some(vq.base.with_row_id),
                     bypass_vector_index: Some(!vq.use_index),
                     full_text_query,
-                    version: None,
+                    ..Default::default()
                 })
             }
             AnyQuery::Query(q) => {
@@ -2225,7 +2419,10 @@ impl NativeTable {
 
                 let columns = match &q.select {
                     Select::All => None,
-                    Select::Columns(cols) => Some(cols.clone()),
+                    Select::Columns(cols) => Some(Box::new(QueryTableRequestColumns {
+                        column_names: Some(cols.clone()),
+                        column_aliases: None,
+                    })),
                     Select::Dynamic(_) => {
                         return Err(Error::NotSupported {
                             message: "Dynamic columns are not supported for server-side query"
@@ -2264,18 +2461,11 @@ impl NativeTable {
                     columns,
                     prefilter: Some(q.prefilter),
                     offset: q.offset.map(|o| o as i32),
-                    ef: None,
-                    refine_factor: None,
-                    distance_type: None,
-                    nprobes: None,
                     vector_column: None, // No vector column for plain queries
                     with_row_id: Some(q.with_row_id),
                     bypass_vector_index: Some(true), // No vector index for plain queries
                     full_text_query,
-                    version: None,
-                    fast_search: None,
-                    lower_bound: None,
-                    upper_bound: None,
+                    ..Default::default()
                 })
             }
         }
@@ -2880,11 +3070,8 @@ impl BaseTable for NativeTable {
 
     /// Delete rows from the table
     async fn delete(&self, predicate: &str) -> Result<DeleteResult> {
-        let mut dataset = self.dataset.get_mut().await?;
-        dataset.delete(predicate).await?;
-        Ok(DeleteResult {
-            version: dataset.version().version,
-        })
+        // Delegate to the submodule implementation
+        delete::execute_delete(self, predicate).await
     }
 
     async fn tags(&self) -> Result<Box<dyn Tags + '_>> {
@@ -3028,8 +3215,8 @@ impl BaseTable for NativeTable {
         Ok(results.into_iter().flatten().collect())
     }
 
-    fn dataset_uri(&self) -> &str {
-        self.uri.as_str()
+    async fn uri(&self) -> Result<String> {
+        Ok(self.uri.clone())
     }
 
     async fn storage_options(&self) -> Option<HashMap<String, String>> {
@@ -3037,7 +3224,7 @@ impl BaseTable for NativeTable {
             .get()
             .await
             .ok()
-            .and_then(|dataset| dataset.storage_options().cloned())
+            .and_then(|dataset| dataset.initial_storage_options().cloned())
     }
 
     async fn index_stats(&self, index_name: &str) -> Result<Option<IndexStatistics>> {
@@ -3202,7 +3389,6 @@ pub struct FragmentSummaryStats {
 #[cfg(test)]
 #[allow(deprecated)]
 mod tests {
-    use std::iter;
     use std::sync::atomic::{AtomicBool, Ordering};
     use std::sync::Arc;
     use std::time::Duration;
@@ -3819,7 +4005,7 @@ mod tests {
                 schema.clone(),
                 vec![
                     Arc::new(Int32Array::from_iter_values(offset..(offset + 10))),
-                    Arc::new(Int32Array::from_iter_values(iter::repeat(age).take(10))),
+                    Arc::new(Int32Array::from_iter_values(std::iter::repeat_n(age, 10))),
                 ],
             )],
             schema,
@@ -4950,7 +5136,13 @@ mod tests {
         assert_eq!(ns_request.k, 10);
         assert_eq!(ns_request.offset, Some(5));
         assert_eq!(ns_request.filter, Some("id > 0".to_string()));
-        assert_eq!(ns_request.columns, Some(vec!["id".to_string()]));
+        assert_eq!(
+            ns_request
+                .columns
+                .as_ref()
+                .and_then(|c| c.column_names.as_ref()),
+            Some(&vec!["id".to_string()])
+        );
         assert_eq!(ns_request.vector_column, Some("vector".to_string()));
         assert_eq!(ns_request.distance_type, Some("l2".to_string()));
         assert!(ns_request.vector.single_vector.is_some());
@@ -4991,7 +5183,13 @@ mod tests {
         assert_eq!(ns_request.k, 20);
         assert_eq!(ns_request.offset, Some(5));
         assert_eq!(ns_request.filter, Some("id > 5".to_string()));
-        assert_eq!(ns_request.columns, Some(vec!["id".to_string()]));
+        assert_eq!(
+            ns_request
+                .columns
+                .as_ref()
+                .and_then(|c| c.column_names.as_ref()),
+            Some(&vec!["id".to_string()])
+        );
         assert_eq!(ns_request.with_row_id, Some(true));
         assert_eq!(ns_request.bypass_vector_index, Some(true));
         assert!(ns_request.vector_column.is_none()); // No vector column for plain queries

rust/lancedb/src/table/delete.rs

@@ -0,0 +1,161 @@
+// SPDX-License-Identifier: Apache-2.0
+// SPDX-FileCopyrightText: Copyright The LanceDB Authors
+use serde::{Deserialize, Serialize};
+
+use super::NativeTable;
+use crate::Result;
+
+#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize, Default)]
+pub struct DeleteResult {
+    // The commit version associated with the operation.
+    // A version of `0` indicates compatibility with legacy servers that do not return
+    /// a commit version.
+    #[serde(default)]
+    pub version: u64,
+}
+
+/// Internal implementation of the delete logic
+///
+/// This logic was moved from NativeTable::delete to keep table.rs clean.
+pub(crate) async fn execute_delete(table: &NativeTable, predicate: &str) -> Result<DeleteResult> {
+    // We access the dataset from the table. Since this is in the same module hierarchy (super),
+    // and 'dataset' is pub(crate), we can access it.
+    let mut dataset = table.dataset.get_mut().await?;
+
+    // Perform the actual delete on the Lance dataset
+    dataset.delete(predicate).await?;
+
+    // Return the result with the new version
+    Ok(DeleteResult {
+        version: dataset.version().version,
+    })
+}
+
+#[cfg(test)]
+mod tests {
+    use crate::connect;
+    use arrow_array::{record_batch, Int32Array, RecordBatch, RecordBatchIterator};
+    use arrow_schema::{DataType, Field, Schema};
+    use std::sync::Arc;
+
+    use crate::query::ExecutableQuery;
+    use futures::TryStreamExt;
+    #[tokio::test]
+    async fn test_delete_simple() {
+        let conn = connect("memory://").execute().await.unwrap();
+
+        // 1. Create a table with values 0 to 9
+        let schema = Arc::new(Schema::new(vec![Field::new("i", DataType::Int32, false)]));
+        let batch = RecordBatch::try_new(
+            schema.clone(),
+            vec![Arc::new(Int32Array::from_iter_values(0..10))],
+        )
+        .unwrap();
+
+        let table = conn
+            .create_table(
+                "test_delete",
+                RecordBatchIterator::new(vec![Ok(batch)], schema),
+            )
+            .execute()
+            .await
+            .unwrap();
+
+        // 2. Verify initial state
+        assert_eq!(table.count_rows(None).await.unwrap(), 10);
+
+        // 3. Execute Delete (removes values > 5)
+        table.delete("i > 5").await.unwrap();
+
+        // 4. Verify results
+        assert_eq!(table.count_rows(None).await.unwrap(), 6); // 0, 1, 2, 3, 4, 5 remain
+
+        // 5. Verify specific data consistency
+        let batches = table
+            .query()
+            .execute()
+            .await
+            .unwrap()
+            .try_collect::<Vec<_>>()
+            .await
+            .unwrap();
+        let batch = &batches[0];
+        let array = batch
+            .column(0)
+            .as_any()
+            .downcast_ref::<Int32Array>()
+            .unwrap();
+
+        // Ensure no value > 5 exists
+        for val in array.iter() {
+            assert!(val.unwrap() <= 5);
+        }
+    }
+    #[tokio::test]
+    async fn rows_removed_schema_same() {
+        let conn = connect("memory://").execute().await.unwrap();
+        let batch = record_batch!(
+            ("id", Int32, [1, 2, 3, 4, 5]),
+            ("name", Utf8, ["a", "b", "c", "d", "e"])
+        )
+        .unwrap();
+        let original_schema = batch.schema();
+
+        let table = conn
+            .create_table(
+                "test_delete_all",
+                RecordBatchIterator::new(vec![Ok(batch)], original_schema.clone()),
+            )
+            .execute()
+            .await
+            .unwrap();
+
+        table.delete("true").await.unwrap();
+
+        assert_eq!(table.count_rows(None).await.unwrap(), 0);
+
+        let current_schema = table.schema().await.unwrap();
+        //check if the original schema is the same as current
+        assert_eq!(current_schema, original_schema);
+    }
+
+    #[tokio::test]
+    async fn test_delete_false_increments_version() {
+        let conn = connect("memory://").execute().await.unwrap();
+
+        // Create a table with 5 rows
+        let batch = record_batch!(("id", Int32, [1, 2, 3, 4, 5])).unwrap();
+
+        let schema = batch.schema();
+
+        let table = conn
+            .create_table(
+                "test_delete_noop",
+                RecordBatchIterator::new(vec![Ok(batch)], schema),
+            )
+            .execute()
+            .await
+            .unwrap();
+
+        // Capture the initial state (Rows = 5, Version = 1)
+        let initial_rows = table.count_rows(None).await.unwrap();
+        let initial_version = table.version().await.unwrap();
+
+        assert_eq!(initial_rows, 5);
+        table.delete("false").await.unwrap();
+
+        // Rows should still be 5
+        let current_rows = table.count_rows(None).await.unwrap();
+        assert_eq!(
+            current_rows, initial_rows,
+            "Data should not change when predicate is false"
+        );
+
+        // version check
+        let current_version = table.version().await.unwrap();
+        assert!(
+            current_version > initial_version,
+            "Table version must increment after delete operation"
+        );
+    }
+}

rust/lancedb/src/test_utils/connection.rs

@@ -5,16 +5,19 @@
 
 use regex::Regex;
 use std::env;
-use std::io::{BufRead, BufReader};
-use std::process::{Child, ChildStdout, Command, Stdio};
+use std::process::Stdio;
+use tokio::io::{AsyncBufReadExt, BufReader};
+use tokio::process::{Child, ChildStdout, Command};
+use tokio::sync::mpsc;
 
 use crate::{connect, Connection};
-use anyhow::{bail, Result};
+use anyhow::{anyhow, bail, Result};
 use tempfile::{tempdir, TempDir};
 
 pub struct TestConnection {
     pub uri: String,
     pub connection: Connection,
+    pub is_remote: bool,
     _temp_dir: Option<TempDir>,
     _process: Option<TestProcess>,
 }
@@ -37,6 +40,56 @@ pub async fn new_test_connection() -> Result<TestConnection> {
     }
 }
 
+async fn spawn_stdout_reader(
+    mut stdout: BufReader<ChildStdout>,
+    port_sender: mpsc::Sender<anyhow::Result<String>>,
+) -> tokio::task::JoinHandle<()> {
+    let print_stdout = env::var("PRINT_LANCEDB_TEST_CONNECTION_SCRIPT_OUTPUT").is_ok();
+    tokio::spawn(async move {
+        let mut line = String::new();
+        let re = Regex::new(r"Query node now listening on 0.0.0.0:(.*)").unwrap();
+        loop {
+            line.clear();
+            let result = stdout.read_line(&mut line).await;
+            if let Err(err) = result {
+                port_sender
+                    .send(Err(anyhow!(
+                        "error while reading from process output: {}",
+                        err
+                    )))
+                    .await
+                    .unwrap();
+                return;
+            } else if result.unwrap() == 0 {
+                port_sender
+                    .send(Err(anyhow!(
+                        " hit EOF before reading port from process output."
+                    )))
+                    .await
+                    .unwrap();
+                return;
+            }
+            if re.is_match(&line) {
+                let caps = re.captures(&line).unwrap();
+                port_sender.send(Ok(caps[1].to_string())).await.unwrap();
+                break;
+            }
+        }
+        loop {
+            line.clear();
+            match stdout.read_line(&mut line).await {
+                Err(_) => return,
+                Ok(0) => return,
+                Ok(_size) => {
+                    if print_stdout {
+                        print!("{}", line);
+                    }
+                }
+            }
+        }
+    })
+}
+
 async fn new_remote_connection(script_path: &str) -> Result<TestConnection> {
     let temp_dir = tempdir()?;
     let data_path = temp_dir.path().to_str().unwrap().to_string();
@@ -57,38 +110,25 @@ async fn new_remote_connection(script_path: &str) -> Result<TestConnection> {
         child: child_result.unwrap(),
     };
     let stdout = BufReader::new(process.child.stdout.take().unwrap());
-    let port = read_process_port(stdout)?;
+    let (port_sender, mut port_receiver) = mpsc::channel(5);
+    let _reader = spawn_stdout_reader(stdout, port_sender).await;
+    let port = match port_receiver.recv().await {
+        None => bail!("Unable to determine the port number used by the phalanx process we spawned, because the reader thread was closed too soon."),
+        Some(Err(err)) => bail!("Unable to determine the port number used by the phalanx process we spawned, because of an error, {}", err),
+        Some(Ok(port)) => port,
+    };
     let uri = "db://test";
     let host_override = format!("http://localhost:{}", port);
     let connection = create_new_connection(uri, &host_override).await?;
     Ok(TestConnection {
         uri: uri.to_string(),
         connection,
+        is_remote: true,
         _temp_dir: Some(temp_dir),
         _process: Some(process),
     })
 }
 
-fn read_process_port(mut stdout: BufReader<ChildStdout>) -> Result<String> {
-    let mut line = String::new();
-    let re = Regex::new(r"Query node now listening on 0.0.0.0:(.*)").unwrap();
-    loop {
-        let result = stdout.read_line(&mut line);
-        if let Err(err) = result {
-            bail!(format!(
-                "read_process_port: error while reading from process output: {}",
-                err
-            ));
-        } else if result.unwrap() == 0 {
-            bail!("read_process_port: hit EOF before reading port from process output.");
-        }
-        if re.is_match(&line) {
-            let caps = re.captures(&line).unwrap();
-            return Ok(caps[1].to_string());
-        }
-    }
-}
-
 #[cfg(feature = "remote")]
 async fn create_new_connection(uri: &str, host_override: &str) -> crate::error::Result<Connection> {
     connect(uri)
@@ -114,6 +154,7 @@ async fn new_local_connection() -> Result<TestConnection> {
     Ok(TestConnection {
         uri: uri.to_string(),
         connection,
+        is_remote: false,
         _temp_dir: Some(temp_dir),
         _process: None,
     })

rust/lancedb/tests/embedding_registry_test.rs

@@ -4,7 +4,6 @@
 use std::{
     borrow::Cow,
     collections::{HashMap, HashSet},
-    iter::repeat,
     sync::Arc,
 };
 
@@ -268,9 +267,10 @@ fn create_some_records() -> Result<impl IntoArrow> {
             schema.clone(),
             vec![
                 Arc::new(Int32Array::from_iter_values(0..TOTAL as i32)),
-                Arc::new(StringArray::from_iter(
-                    repeat(Some("hello world".to_string())).take(TOTAL),
-                )),
+                Arc::new(StringArray::from_iter(std::iter::repeat_n(
+                    Some("hello world".to_string()),
+                    TOTAL,
+                ))),
             ],
         )
         .unwrap()]

rust/lancedb/tests/embeddings_parallel_test.rs

@@ -0,0 +1,253 @@
+// SPDX-License-Identifier: Apache-2.0
+// SPDX-FileCopyrightText: Copyright The LanceDB Authors
+
+use std::{
+    borrow::Cow,
+    sync::{
+        atomic::{AtomicUsize, Ordering},
+        Arc,
+    },
+    time::Duration,
+};
+
+use arrow::buffer::NullBuffer;
+use arrow_array::{
+    Array, FixedSizeListArray, Float32Array, RecordBatch, RecordBatchIterator, StringArray,
+};
+use arrow_schema::{DataType, Field, Schema};
+use lancedb::{
+    embeddings::{EmbeddingDefinition, EmbeddingFunction, MaybeEmbedded, WithEmbeddings},
+    Error, Result,
+};
+
+#[derive(Debug)]
+struct SlowMockEmbed {
+    name: String,
+    dim: usize,
+    delay_ms: u64,
+    call_count: Arc<AtomicUsize>,
+}
+
+impl SlowMockEmbed {
+    pub fn new(name: String, dim: usize, delay_ms: u64) -> Self {
+        Self {
+            name,
+            dim,
+            delay_ms,
+            call_count: Arc::new(AtomicUsize::new(0)),
+        }
+    }
+
+    pub fn get_call_count(&self) -> usize {
+        self.call_count.load(Ordering::SeqCst)
+    }
+}
+
+impl EmbeddingFunction for SlowMockEmbed {
+    fn name(&self) -> &str {
+        &self.name
+    }
+
+    fn source_type(&self) -> Result<Cow<'_, DataType>> {
+        Ok(Cow::Owned(DataType::Utf8))
+    }
+
+    fn dest_type(&self) -> Result<Cow<'_, DataType>> {
+        Ok(Cow::Owned(DataType::new_fixed_size_list(
+            DataType::Float32,
+            self.dim as _,
+            true,
+        )))
+    }
+
+    fn compute_source_embeddings(&self, source: Arc<dyn Array>) -> Result<Arc<dyn Array>> {
+        // Simulate slow embedding computation
+        std::thread::sleep(Duration::from_millis(self.delay_ms));
+        self.call_count.fetch_add(1, Ordering::SeqCst);
+
+        let len = source.len();
+        let inner = Arc::new(Float32Array::from(vec![Some(1.0); len * self.dim]));
+        let field = Field::new("item", inner.data_type().clone(), false);
+        let arr = FixedSizeListArray::new(
+            Arc::new(field),
+            self.dim as _,
+            inner,
+            Some(NullBuffer::new_valid(len)),
+        );
+
+        Ok(Arc::new(arr))
+    }
+
+    fn compute_query_embeddings(&self, _input: Arc<dyn Array>) -> Result<Arc<dyn Array>> {
+        unimplemented!()
+    }
+}
+
+fn create_test_batch() -> Result<RecordBatch> {
+    let schema = Arc::new(Schema::new(vec![Field::new("text", DataType::Utf8, false)]));
+    let text = StringArray::from(vec!["hello", "world"]);
+    RecordBatch::try_new(schema, vec![Arc::new(text)]).map_err(|e| Error::Runtime {
+        message: format!("Failed to create test batch: {}", e),
+    })
+}
+
+#[test]
+fn test_single_embedding_fast_path() {
+    // Single embedding should execute without spawning threads
+    let batch = create_test_batch().unwrap();
+    let schema = batch.schema();
+
+    let embed = Arc::new(SlowMockEmbed::new("test".to_string(), 2, 10));
+    let embedding_def = EmbeddingDefinition::new("text", "test", Some("embedding"));
+
+    let reader = RecordBatchIterator::new(vec![Ok(batch)], schema);
+    let embeddings = vec![(embedding_def, embed.clone() as Arc<dyn EmbeddingFunction>)];
+    let mut with_embeddings = WithEmbeddings::new(reader, embeddings);
+
+    let result = with_embeddings.next().unwrap().unwrap();
+    assert!(result.column_by_name("embedding").is_some());
+    assert_eq!(embed.get_call_count(), 1);
+}
+
+#[test]
+fn test_multiple_embeddings_parallel() {
+    // Multiple embeddings should execute in parallel
+    let batch = create_test_batch().unwrap();
+    let schema = batch.schema();
+
+    let embed1 = Arc::new(SlowMockEmbed::new("embed1".to_string(), 2, 100));
+    let embed2 = Arc::new(SlowMockEmbed::new("embed2".to_string(), 3, 100));
+    let embed3 = Arc::new(SlowMockEmbed::new("embed3".to_string(), 4, 100));
+
+    let def1 = EmbeddingDefinition::new("text", "embed1", Some("emb1"));
+    let def2 = EmbeddingDefinition::new("text", "embed2", Some("emb2"));
+    let def3 = EmbeddingDefinition::new("text", "embed3", Some("emb3"));
+
+    let reader = RecordBatchIterator::new(vec![Ok(batch)], schema);
+    let embeddings = vec![
+        (def1, embed1.clone() as Arc<dyn EmbeddingFunction>),
+        (def2, embed2.clone() as Arc<dyn EmbeddingFunction>),
+        (def3, embed3.clone() as Arc<dyn EmbeddingFunction>),
+    ];
+    let mut with_embeddings = WithEmbeddings::new(reader, embeddings);
+
+    let result = with_embeddings.next().unwrap().unwrap();
+
+    // Verify all embedding columns are present
+    assert!(result.column_by_name("emb1").is_some());
+    assert!(result.column_by_name("emb2").is_some());
+    assert!(result.column_by_name("emb3").is_some());
+
+    // Verify all embeddings were computed
+    assert_eq!(embed1.get_call_count(), 1);
+    assert_eq!(embed2.get_call_count(), 1);
+    assert_eq!(embed3.get_call_count(), 1);
+}
+
+#[test]
+fn test_embedding_column_order_preserved() {
+    // Verify that embedding columns are added in the same order as definitions
+    let batch = create_test_batch().unwrap();
+    let schema = batch.schema();
+
+    let embed1 = Arc::new(SlowMockEmbed::new("embed1".to_string(), 2, 10));
+    let embed2 = Arc::new(SlowMockEmbed::new("embed2".to_string(), 3, 10));
+    let embed3 = Arc::new(SlowMockEmbed::new("embed3".to_string(), 4, 10));
+
+    let def1 = EmbeddingDefinition::new("text", "embed1", Some("first"));
+    let def2 = EmbeddingDefinition::new("text", "embed2", Some("second"));
+    let def3 = EmbeddingDefinition::new("text", "embed3", Some("third"));
+
+    let reader = RecordBatchIterator::new(vec![Ok(batch)], schema);
+    let embeddings = vec![
+        (def1, embed1 as Arc<dyn EmbeddingFunction>),
+        (def2, embed2 as Arc<dyn EmbeddingFunction>),
+        (def3, embed3 as Arc<dyn EmbeddingFunction>),
+    ];
+    let mut with_embeddings = WithEmbeddings::new(reader, embeddings);
+
+    let result = with_embeddings.next().unwrap().unwrap();
+    let result_schema = result.schema();
+
+    // Original column is first
+    assert_eq!(result_schema.field(0).name(), "text");
+    // Embedding columns follow in order
+    assert_eq!(result_schema.field(1).name(), "first");
+    assert_eq!(result_schema.field(2).name(), "second");
+    assert_eq!(result_schema.field(3).name(), "third");
+}
+
+#[test]
+fn test_embedding_error_propagation() {
+    // Test that errors from embedding computation are properly propagated
+    #[derive(Debug)]
+    struct FailingEmbed {
+        name: String,
+    }
+
+    impl EmbeddingFunction for FailingEmbed {
+        fn name(&self) -> &str {
+            &self.name
+        }
+
+        fn source_type(&self) -> Result<Cow<'_, DataType>> {
+            Ok(Cow::Owned(DataType::Utf8))
+        }
+
+        fn dest_type(&self) -> Result<Cow<'_, DataType>> {
+            Ok(Cow::Owned(DataType::new_fixed_size_list(
+                DataType::Float32,
+                2,
+                true,
+            )))
+        }
+
+        fn compute_source_embeddings(&self, _source: Arc<dyn Array>) -> Result<Arc<dyn Array>> {
+            Err(Error::Runtime {
+                message: "Intentional failure".to_string(),
+            })
+        }
+
+        fn compute_query_embeddings(&self, _input: Arc<dyn Array>) -> Result<Arc<dyn Array>> {
+            unimplemented!()
+        }
+    }
+
+    let batch = create_test_batch().unwrap();
+    let schema = batch.schema();
+
+    let embed = Arc::new(FailingEmbed {
+        name: "failing".to_string(),
+    });
+    let def = EmbeddingDefinition::new("text", "failing", Some("emb"));
+
+    let reader = RecordBatchIterator::new(vec![Ok(batch)], schema);
+    let embeddings = vec![(def, embed as Arc<dyn EmbeddingFunction>)];
+    let mut with_embeddings = WithEmbeddings::new(reader, embeddings);
+
+    let result = with_embeddings.next().unwrap();
+    assert!(result.is_err());
+    let err_msg = format!("{}", result.err().unwrap());
+    assert!(err_msg.contains("Intentional failure"));
+}
+
+#[test]
+fn test_maybe_embedded_with_no_embeddings() {
+    // Test that MaybeEmbedded::No variant works correctly
+    let batch = create_test_batch().unwrap();
+    let schema = batch.schema();
+
+    let reader = RecordBatchIterator::new(vec![Ok(batch.clone())], schema.clone());
+    let table_def = lancedb::table::TableDefinition {
+        schema: schema.clone(),
+        column_definitions: vec![lancedb::table::ColumnDefinition {
+            kind: lancedb::table::ColumnKind::Physical,
+        }],
+    };
+
+    let mut maybe_embedded = MaybeEmbedded::try_new(reader, table_def, None).unwrap();
+
+    let result = maybe_embedded.next().unwrap().unwrap();
+    assert_eq!(result.num_columns(), 1);
+    assert_eq!(result.column(0).as_ref(), batch.column(0).as_ref());
+}