ai-workspace-services/litellm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(audio): detect gpt-realtime models as audio-capable for Chat Completions

Browse Source 
gpt-realtime-1.5 supports Chat Completions with audio params but
is_model_gpt_audio_model only checked for "audio" in the model name.
Add "realtime" check so the audio parameter is passed through correctly.
This commit is contained in:
Chesars 2026-02-27 16:06:10 -03:00
parent 1f887547f6
commit c7ab631bb0
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
litellm/llms/openai/chat/gpt_audio_transformation.py
View File

@ -29,7 +29,9 @@ class OpenAIGPTAudioConfig(OpenAIGPTConfig):
return all_openai_params + audio_specific_params
def is_model_gpt_audio_model(self, model: str) -> bool:
if model in litellm.open_ai_chat_completion_models and "audio" in model:
if model in litellm.open_ai_chat_completion_models and (
"audio" in model or "realtime" in model
):
return True
return False