Feature Request : compatibility with LiteLLM #148
Description
Due diligence
- I have done my due diligence in trying to find the answer myself.
Topic
Other / All
Question
Description
It would be great to add LiteLLM’s audio transcription API integration to this project.
This would allow users to easily transcribe audio streams using a unified and lightweight interface.
LiteLLM provides a simple wrapper around various LLM and ASR providers, enabling flexible and modular integration.
Their documentation can be found here:
🔗 LiteLLM Audio Transcription Docs
Motivation
Currently, the project focuses on delayed stream modeling, but does not directly support external speech-to-text APIs.
Integrating LiteLLM would:
- Simplify audio transcription setup
- Allow multiple backend providers (e.g. OpenAI, Anthropic, Whisper)
- Enable easy experimentation with different STT models
- Support asynchronous and streaming processing for real-time use cases