diff --git a/sbv2_core/src/main.rs b/sbv2_core/src/main.rs
index 925e735..b84e44a 100644
--- a/sbv2_core/src/main.rs
+++ b/sbv2_core/src/main.rs
@@ -24,8 +24,7 @@ fn main() -> anyhow::Result<()> {
         tones.clone(),
         lang_ids.clone(),
         style_vector.clone(),
-        0.0,
-        0.5,
+        tts::SynthesizeOptions::default(),
     )?;
     std::fs::write("output.wav", data)?;
     let now = Instant::now();
@@ -45,8 +44,7 @@ fn main() -> anyhow::Result<()> {
             tones.clone(),
             lang_ids.clone(),
             style_vector.clone(),
-            0.0,
-            1.0,
+            tts::SynthesizeOptions::default(),
         )?;
     }
     println!(
diff --git a/sbv2_core/src/tts.rs b/sbv2_core/src/tts.rs
index 0d50096..9ff3197 100644
--- a/sbv2_core/src/tts.rs
+++ b/sbv2_core/src/tts.rs
@@ -206,8 +206,7 @@ impl TTSModelHolder {
         tones: Array1<i64>,
         lang_ids: Array1<i64>,
         style_vector: Array1<f32>,
-        sdp_ratio: f32,
-        length_scale: f32,
+        options: SynthesizeOptions,
     ) -> Result<Vec<u8>> {
         let buffer = model::synthesize(
             &self.find_model(ident)?.vits2,
@@ -216,9 +215,25 @@ impl TTSModelHolder {
             tones,
             lang_ids,
             style_vector,
-            sdp_ratio,
-            length_scale,
+            options.sdp_ratio,
+            options.length_scale,
         )?;
         Ok(buffer)
     }
 }
+
+pub struct SynthesizeOptions {
+    sdp_ratio: f32,
+    length_scale: f32,
+    split_sentences: bool,
+}
+
+impl Default for SynthesizeOptions {
+    fn default() -> Self {
+        SynthesizeOptions {
+            sdp_ratio: 0.0,
+            length_scale: 1.0,
+            split_sentences: true,
+        }
+    }
+}