Developer Benjamin Schoepke released an open‑source MCP server that lets AI assistants command Ableton Live through Python evaluation. By exposing the full Ableton object model, the bridge can perform any action the DAW supports, from MIDI manipulation to plugin control. The project aims to let users issue voice or code prompts while hands are otherwise occupied while coding.

Installation requires only a few commands; users can instruct any LLM—Codex, Claude, Gemini, or Copilot—to pull the repository and launch the server on macOS or Windows. Schoepke tested it on Ableton Live Suite 12.3.8 under macOS Ventura, noting that the MCP edits Live sets directly and may corrupt files if backups are missing. A short demo video showcases a full‑song generation workflow in real time.

Beyond voice control, the bridge can drive external synthesizers, load VSTs such as Serum, and trigger VJ plugins like VideoSync for live visualisation. Users report that pre‑defined tool shortcuts speed common tasks, while the open Python layer retains full flexibility. The repo now serves as a practical template for developers aiming to merge AI instruction with real‑time music production.