kanaworksai-hue/audio-studio-mcp is a local sound studio that creates background music and sound effects from natural-language requests. The latest publicly released version is v0.2.1. It can be used in MCP clients (the protocol through which AI tools communicate with external tools) such as Claude Code, Claude Desktop, and Cursor, and runs on a Mac or Windows PC on top of Stable Audio 3.
Generation, editing, and export all finish locally. Natural-language requests produce background music, instrument loops, and sound effects, and the resulting audio can be polished with looping, fades, and volume adjustment before being exported in WAV, OGG, or FLAC format. It uses MLX on Apple Silicon and PyTorch on NVIDIA GPUs under Windows, and needs no cloud service or API key. It is written in Python, licensed under MIT, and has 374 GitHub stars. The repository was created on October 18, 2026.
The scope that can be judged, however, is limited. The given material contains no changelog for v0.2.1, so what this release changed cannot be confirmed. The README claims that results match those of the official command-line tool sa3 with the same seed value, but no material comparing generation speed and audio quality against other tools is presented. The research record gives quality 7.0, novelty 6.0, technology 6.0, and momentum 8.5 on a 10-point scale. Audio quality and response speed can only be known from the README's claims, and the output has to be produced directly before they can be judged.