Generates high-fidelity music and sound effects from text descriptions using Meta's AudioCraft framework.
This skill enables Claude to implement and orchestrate Meta's AudioCraft suite, including MusicGen for text-to-music and AudioGen for sound effect creation. It provides comprehensive patterns for building audio-centric applications, configuring model variants from small (300M) to large (3.3B), and utilizing advanced features like melody-conditioned generation, stereo output, and neural audio compression with EnCodec. This is an essential toolkit for developers and researchers building creative tools, game sound engines, or multimedia processing pipelines that require controllable, state-of-the-art audio synthesis.
주요 기능
01Style-based conditioning and reference-based music synthesis
02Environmental sound effect synthesis through AudioGen
03High-fidelity neural audio compression and reconstruction via EnCodec
04Text-to-music generation using MusicGen with custom duration and sampling
05Melody-conditioned generation for rhythmic and pitch-based control
06384 GitHub stars
사용 사례
01Generating custom, royalty-free background music for video production and games
02Building AI-powered music production tools with style transfer and melody guidance
03Creating realistic environmental foley and sound effects for immersive applications
