Merge pull request #3697 from pipecat-ai/mb/soniox-rt-4

Update SonioxSTTService default model to stt-rt-v4

changelog/3697.changed.2.md

@@ -0,0 +1 @@
+- Update `SonioxSTTService` to set `vad_force_turn_endpoint` to `True`. This setting disabled the turn detection logic available natively in Soniox. Instead, Soniox relies on a local VAD to finalize the transcript. This configuration meaningfully reduces the time to final segment for Soniox. With this setting enabled, Soniox outputs a transcript in ~250ms (median). Pipecat enables smart-turn detection by default using the `LocalSmartTurnAnalyzerV3`. To use the native turn detection logic in Soniox, just set `vad_force_turn_endpoint` to `False`.

changelog/3697.changed.md

@@ -0,0 +1 @@
+- Update `SonioxSTTService` default model to `stt-rt-v4`.

examples/foundational/07za-interruptible-soniox.py

@@ -24,7 +24,8 @@ from pipecat.runner.types import RunnerArguments
 from pipecat.runner.utils import create_transport
 from pipecat.services.cartesia.tts import CartesiaTTSService
 from pipecat.services.openai.llm import OpenAILLMService
-from pipecat.services.soniox.stt import SonioxSTTService
+from pipecat.services.soniox.stt import SonioxInputParams, SonioxSTTService
+from pipecat.transcriptions.language import Language
 from pipecat.transports.base_transport import BaseTransport, TransportParams
 from pipecat.transports.daily.transport import DailyParams
 from pipecat.transports.websocket.fastapi import FastAPIWebsocketParams
@@ -52,6 +53,10 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
 
     stt = SonioxSTTService(
         api_key=os.getenv("SONIOX_API_KEY"),
+        params=SonioxInputParams(
+            language_hints=[Language.EN],
+            language_hints_strict=True,
+        ),
     )
 
     tts = CartesiaTTSService(

src/pipecat/services/soniox/stt.py

@@ -93,7 +93,7 @@ class SonioxInputParams(BaseModel):
         client_reference_id: Client reference ID to use for transcription.
     """
 
-    model: str = "stt-rt-preview"
+    model: str = "stt-rt-v4"
 
     audio_format: Optional[str] = "pcm_s16le"
     num_channels: Optional[int] = 1
@@ -152,7 +152,7 @@ class SonioxSTTService(WebsocketSTTService):
         url: str = "wss://stt-rt.soniox.com/transcribe-websocket",
         sample_rate: Optional[int] = None,
         params: Optional[SonioxInputParams] = None,
-        vad_force_turn_endpoint: bool = False,
+        vad_force_turn_endpoint: bool = True,
         ttfs_p99_latency: Optional[float] = SONIOX_TTFS_P99,
         **kwargs,
     ):
@@ -164,7 +164,8 @@ class SonioxSTTService(WebsocketSTTService):
             sample_rate: Audio sample rate.
             params: Additional configuration parameters, such as language hints, context and
                 speaker diarization.
-            vad_force_turn_endpoint: Listen to `VADUserStoppedSpeakingFrame` to send finalize message to Soniox. If disabled, Soniox will detect the end of the speech.
+            vad_force_turn_endpoint: Listen to `VADUserStoppedSpeakingFrame` to send finalize message to Soniox.
+                If disabled, Soniox will detect the end of the speech. Defaults to True.
             ttfs_p99_latency: P99 latency from speech end to final transcript in seconds.
                 Override for your deployment. See https://github.com/pipecat-ai/stt-benchmark
             **kwargs: Additional arguments passed to the STTService.