Generate game-ready sprites, animations, 3D models, video, music and sound effects from your AI assistant through Ludo.ai's hosted MCP server.
Audio Studio MCP
by KanaWorks_AI
Released
MIT
Summary
Generate music loops and sound effects locally with Stable Audio 3, then loop, normalize and export them, from any MCP client.
Features
- Text to background music, instrument loops and sound effects
- Edit tools: variation, inpaint a time range, or extend a track
- Seamless loops, fades and LUFS loudness normalization
- Exports 48 kHz WAV, OGG or FLAC with a reproducible JSON sidecar
- Runs fully offline on Apple Silicon (MLX) or NVIDIA GPUs (PyTorch)
Installation
Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.
Description
Audio Studio MCP turns Claude Code, Claude Desktop, Cursor or any other MCP client into a local sound studio. Describe background music, an instrument loop or a sound effect in plain language and the server generates one to four candidates with Stable Audio 3 on your own machine, with no cloud service and no API key.
Three tools cover the workflow. generate_audio creates music or effects, edit_audio re-styles a clip, regenerates a time range or extends a track, and process_audio finishes the result with silence trimming, loudness normalization, seamless loops, fades and export to 48 kHz WAV, OGG or FLAC. Every file gets a JSON sidecar recording its prompt, seed and settings, so results are reproducible, and the server teaches the client the official Stable Audio 3 prompt format.
It runs on Apple Silicon Macs through MLX and on Windows PCs with an NVIDIA GPU through PyTorch. The installer needs git and ffmpeg and downloads the medium model (about 5.9 GB on macOS, about 10.4 GB on Windows). It suits game developers and video editors who want royalty-friendly placeholder audio without leaving their editor. MIT licensed.
Related MCP Servers
Red Hat's zero-dependency C++23 code-context engine — ranked call graphs and blast-radius analysis, indexing a repo in under half a second with no server and no database.
Persistent memory for coding agents as a single Go binary — SQLite and FTS5 behind MCP, an HTTP API, a CLI and a TUI, with no Node, Python or Docker required.
Indexes the coding-agent sessions already sitting on your disk — months of history from before you installed it — and serves them back over MCP. 3.5 GB searched in ~1.5 ms, no LLM, no embeddings.