Merge pull request #3155 from pipecat-ai/mb/fix-sarvam-tts-not-flushing

fix: flush audio in SarvamTTSService

CHANGELOG.md

@@ -22,6 +22,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Fixed
 
+- Fixed an issue in `SarvamTTSService` where the last sentence was not being
+  spoken. Now, audio is flushed when the TTS services receives the
+  `LLMFullResponseEndFrame` or `EndFrame`.
+
 - Fixed an issue in `AWSTranscribeSTTService` where the `region` arg was
   always set to `us-east-1` when providing an AWS_REGION env var.
 

src/pipecat/services/sarvam/tts.py

@@ -514,9 +514,11 @@ class SarvamTTSService(InterruptibleTTSService):
 
     async def process_frame(self, frame: Frame, direction: FrameDirection):
         """Process a frame and flush audio if it's the end of a full response."""
-        if isinstance(frame, LLMFullResponseEndFrame):
+        await super().process_frame(frame, direction)
+
+        # When the LLM finishes responding, flush any remaining text in Sarvam's buffer
+        if isinstance(frame, (LLMFullResponseEndFrame, EndFrame)):
             await self.flush_audio()
-        return await super().process_frame(frame, direction)
 
     async def _update_settings(self, settings: Mapping[str, Any]):
         """Update service settings and reconnect if voice changed."""