Revert the default model to be "gpt-4o-realtime-preview-2024-12-17" In OpenAIRealtimeBetaLLMService

CHANGELOG.md

@@ -130,7 +130,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 - Added support in `OpenAIRealtimeBetaLLMService` for a slate of new features:
 
-  - The `'gpt-4o-transcribe-latest'` input audio transcription model.
+  - The `'gpt-4o-transcribe-latest'` input audio transcription model, along
+    with new `language` and `prompt` options specific to that model.
   - The `input_audio_noise_reduction` session property.
 
     ```python
@@ -171,9 +172,6 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Updated the default model for `CartesiaTTSService` and
   `CartesiaHttpTTSService` to `sonic-2`.
 
-- Updated the default model for `OpenAIRealtimeBetaLLMService` to
-  `gpt-4o-realtime-preview-latest`.
-
 - Function calls are now executed in tasks. This means that the pipeline will
   not be blocked while the function call is being executed.
 

src/pipecat/services/openai_realtime_beta/openai.py

@@ -89,8 +89,7 @@ class OpenAIRealtimeBetaLLMService(LLMService):
         self,
         *,
         api_key: str,
-        # model: str = "gpt-4o-realtime-preview-2024-12-17",
-        model: str = "gpt-4o-realtime-preview-latest",
+        model: str = "gpt-4o-realtime-preview-2024-12-17",
         base_url: str = "wss://api.openai.com/v1/realtime",
         session_properties: events.SessionProperties = events.SessionProperties(),
         start_audio_paused: bool = False,