Commit Graph
2 Commits
Author SHA1 Message Date
whit3rabbitandClaude Sonnet 4.6 be038e4164 feat: add provider catalog, bedrock native passthrough, and generic passthrough
Introduces anyllm_providers crate as metadata-only catalog (ProviderDef,
ModelDef, registry). Wires any provider-id as BACKEND via OpenAIClient.
Adds bedrock_native.rs (Converse/InvokeModel with SigV4) and
generic_passthrough.rs catch-all for Translate mode. Updates config,
backend clients, streaming, token counting, and compatibility tests
to support the expanded provider surface. Updates model pricing and docs.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-08 16:33:59 -05:00
whit3rabbitandClaude Opus 4.6 d6b9d61071 feat: add caching, batch API, cost tracking, budget/RBAC, audio/image passthrough
LiteLLM parity features:
- Response caching (in-memory moka, optional Redis tier) with per-request TTL
- Batch API (file upload, job creation/listing via OpenAI delegation)
- Per-key budget enforcement (daily/monthly/lifetime) with lazy period reset
- RBAC (admin/developer key roles, developer keys blocked from /admin/)
- Audio transcription/speech and image generation passthrough
- Fallback chain config (YAML-based, 5xx/429 failover)
- Cost tracking foundation (model pricing DB, per-key spend queries)

Code quality cleanup (simplify pass):
- Extract try_cache_response helper (was copy-pasted 4x)
- Extract common_routes for batch/models (was duplicated across 3 HandlerMode arms)
- Deduplicate embeddings_passthrough to delegate to raw_passthrough
- Remove dead code: inject_cost_header, BudgetDuration::from_str_lossy, duplicate CacheConfig
- Collapse epoch_to_ymd wrapper into pub(crate) days_to_ymd

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-26 20:23:38 -05:00