Voice & media · harness

FunASR

An open-source speech recognition toolkit with an MCP server that lets AI agents transcribe audio, detect voice activity, identify speakers, and recognize emotions. Provides an OpenAI-compatible local API for seamless integration with agent frameworks like LangChain and AutoGen.

Connects to: LangChain, AutoGen, Claude, Cursor, OpenAI-compatible API · · MIT 19,889★

Use it with an AI agent

Loadbay is an MCP server, so an agent can search the catalog and find this harness:

claude mcp add --transport http loadbay https://loadbay.xyz/api/mcp