TAU-HOME.COM
LOADING

Audio Studio MCP Launches Windows and NVIDIA GPU Support for Local AI Audio

Open-source local tool Audio Studio MCP has launched Windows support, enabling cloud-free music and SFX generation directly on NVIDIA GPU workstations.

tau · October 19, 2026

#AI #Audio #MCP #Windows #SoundDesign #LocalAI

Audio Studio MCP Launches Windows and NVIDIA GPU Support for Local AI Audio

The generative audio space is shifting from cloud APIs toward autonomous, on-device synthesis. On October 19, 2026, developer @KanaWorks_AI announced official Windows support for Audio Studio MCP, an open-source local AI sound studio integrating Stable Audio 3 into developer tools.

Local Sound Synthesis and Agent-Centric MCP Architecture

Previously limited to Apple Silicon Macs, Audio Studio MCP now supports Windows PCs with NVIDIA GPUs. Operating as a Model Context Protocol (MCP) server, it connects into AI assistants like Claude Code, Claude Desktop, and Cursor rather than running as a standalone GUI. Powered by Stability AI's Stable Audio 3 PyTorch API, the tool lets assistants handle acoustic parameters directly from conversational user requests. A single prompt can generate BGM, instrument loops, and sound effects across up to four candidate variations, alongside style transfer, regional regeneration, and track extension.

Bar-Locked Looping and Game Engine Mastering Pipelines

Audio Studio MCP includes automated processing tailored for game development and media authoring. Compositions can be sliced to tempo and bar counts—such as an 11-bar loop at 85 BPM—for seamless playback. Outputs receive automated LUFS loudness normalization and fade shaping, exporting in 48 kHz WAV, OGG, or FLAC formats ready for Unity and Godot. Every file includes a companion .json sidecar recording prompt, seed, and parameters, ensuring full reproducibility.

Hardware Benchmarks, Privacy, and Caveats

Author tests on an NVIDIA RTX 3060 12GB with a pre-loaded medium model logged generation times of about 1.3 seconds for a 20-second clip and 15 seconds for a 380-second track, peaking at 5.2 GB VRAM. While on-device execution guarantees privacy without API fees, speeds vary by GPU hardware, and the initial announcement omitted direct setup links, requiring manual repository verification.

Sources

Impact-Site-Verification: 9b3ebd10-6c45-44cf-a014-5b8c19b7b285