AudioCraft
by Meta AI · open-source · Last verified 2026-03-17
AudioCraft is Meta AI's open-source audio generation framework that unifies MusicGen (music), AudioGen (sound effects), and EnCodec (neural audio codec) under a single platform, enabling text-to-audio generation across music, environmental sounds, and audio effects. It provides a complete pipeline for generative audio research and production, and its modular design has made it the dominant open-source reference for audio LLM work.
https://github.com/facebookresearch/audiocraft ↗B
B—Above Average
Adoption: BQuality: AFreshness: BCitations: AEngagement: F
Specifications
- License
- MIT
- Pricing
- open-source
- Capabilities
- text-to-music, text-to-sound-effects, audio-codec, fine-tuning, multi-model-framework
- Integrations
- huggingface, pytorch
- Use Cases
- audio-research, game-audio, film-sound-design, music-generation, environmental-audio
- API Available
- Yes
- Parameters
- 1.5B (MusicGen Large)
- Context Window
- N/A
- Modalities
- text, audio
- Training Cutoff
- 2023
- Tags
- audio-generation, music-generation, sound-effects, open-source, meta-ai
- Added
- 2026-03-17
- Completeness
- 100%
Index Score
62.6Adoption
66
Quality
81
Freshness
65
Citations
80
Engagement
0