Open source alternatives to AudioCraft
200 open source apps like AudioCraft, closest match first.
Top picks
- H
HarmonAI
LaunchCreate music effortlessly with pre-made sound kits, generative audio tools, and receive feedback from the vibrant community.
- W
Wav0
LaunchAI-Native Studio for everything audio. Sound Design/Audio Editing/AI Producer/Fast Exports. Built for web, runs anywhere.
- A
ACE-Step-Studio
LaunchPortable AI music generator — full songs with vocals, covers, music videos. One-click install, 100% offline, NVIDIA GPU.
- Y

yovoice
Launchyovoice is an open source studio for making voices and music on your own computer, available for macOS and Windows. It turns text into speech with control over emotion, clones a voice from a short reference recording, and can design a new v
- R
Remiqora
LaunchFree, open-source local AI music studio: ACE-Step 1.5 and YuE2-3B text-to-music, stem separation, an AI arranger that adds parts to your track, a multitrack DAW, MIDI transcription and LoRA training.
Self-hosted
- A

AI Mastering
LaunchAI Mastering is a web service that masters an audio track automatically using AI. You upload a song or mix, the tool analyzes its loudness and dynamic range, and it applies a limiter and processing to bring the sound closer to commercial re
More like AudioCraft
Generative Media Skills
Research-backed agent skills and tools for premium image, video, audio, voice

claude-ai-music-skills
Claude Code plugin for AI music creation covering lyrics, Suno style prompts
Tegridy-MIDI-Dataset
Tegridy MIDI Dataset for precise and effective Music AI models creation
About AudioCraft
AudioCraft
A single-stop code base for generative audio needs, by Meta
AudioCraft is Meta's open-source code base for generative audio. It bundles three models in one project: MusicGen creates music from text prompts, AudioGen creates sound effects and environmental audio from text, and EnCodec is a neural codec that compresses raw audio into discrete tokens and turns them back into a waveform. MusicGen and AudioGen share one design, where a single autoregressive language model predicts token streams and text conditioning steers the result. The project is for researchers, developers and audio engineers who want to run or train generative audio models on their own machines rather than rely on a hosted app. The code and model cards are published on GitHub, so you can read how each model was built and what it was trained for before you use it. The site presents it as one place for the whole workflow, from compressing audio to generating new music and sound.
- Price
- Open source
- Category
- AI
- Website
- metademolab.com