Give Your AI Agent Ears: OpenAI-Compatible API + MCP Server

[LauraGPT]

We released two features for agent developers:

OpenAI-Compatible API

pip install funasr fastapi uvicorn python-multipart
python examples/openai_api/server.py --device cuda

Then use with OpenAI SDK:

from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="x")
result = client.audio.transcriptions.create(model="sensevoice", file=open("a.wav","rb"))

MCP Server (Claude/Cursor)

Add to your MCP config and AI assistants can transcribe audio directly.

Why FunASR for Agents?

• 170x realtime, self-hosted, no API costs
• Speaker diarization + emotion detection
• 50+ languages

Docs: examples/openai_api/ and examples/mcp_server/