whisper.cpp

High-performance C/C++ port of Whisper for fast local and on-device speech-to-text with no Python runtime.

Connects to: ggml, CUDA, Core ML · C++ · MIT 50,800★

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

• Source: https://github.com/ggml-org/whisper.cpp
• This harness as JSON: /api/harnesses/whisper-cpp
• Agent setup: /setup.md
• Browse all 370+ harnesses on Loadbay