Music and sound models
Score, sound effects and audio repair.
- Catalogued here
- 20 models
- Open weights
- 5
- With an API
- 10
What to know about music and sound
Score, sound effects and audio repair. This is the category where ownership is least settled: several vendors keep the copyright in what you generate, or grant a synchronisation licence rather than ownership, and at least two contradict themselves between their marketing and their terms. Read the commercial-use block on every page here before you cut anything to picture.
Duration and resolution, side by side
Only the models whose vendor publishes a figure appear here. "Not established" means no page we could retrieve stated one.
| Model | Duration | Resolution | Weights | Commercial use |
|---|---|---|---|---|
| Eleven Music (music_v2) | 600 s | Up to 48 kHz; default output mp3_48000_192 for v2 models (mp3_44100_128 for v1); PCM up to 48 kHz and Opus also available. High-fidelity export is limited to Pro, Scale, Business and Enterprise tiers; lower tiers get MP3 44.1 kHz at 128-192 kbps. | Not released | Permitted subject to conditions |
| Stable Audio 3.0 Medium | 380 s | not established | Downloadable | Permitted subject to conditions |
| AIVA | 330 s | MP3 and MIDI on Free and Standard; WAV high quality and all formats on Pro | Not released | The vendor keeps the copyright |
| ACE-Step | 240 s | not established | Downloadable | Vendor states commercial use is permitted |
| Google Lyria 3 Pro | 180 s | 44.1 kHz stereo; MP3 by default, WAV optional | Not released | Permitted subject to conditions |
| Udio v1.5 (Allegro) | 130 s | not established | Not released | Rights are not withheld, but output cannot be exported |
| Stable Audio 3.0 Small and Small SFX | 120 s | not established | Downloadable | Permitted subject to conditions |
| ElevenLabs Sound Effects (eleven_text_to_sound_v2) | 30 s | MP3 from 22.05 kHz/32 kbps up to 44.1 kHz/192 kbps; WAV at 48 kHz for non-looping effects; PCM 8 kHz-48 kHz and Opus 48 kHz also available | Not released | Permitted subject to conditions |
| Google Lyria 3 Clip | 30 s | 44.1 kHz stereo MP3 | Not released | Permitted subject to conditions |
| Stable Audio Open Small | 11 s | 44.1 kHz stereo | Downloadable | Permitted subject to conditions |
| Beatoven.ai maestro | not established | MP3 or WAV (bit depth and sample rate not documented) | Not released | The vendor keeps the copyright |
| Google Lyria RealTime | not established | 48 kHz stereo 16-bit PCM | Not released | Permitted subject to conditions |
| Mubert | not established | High quality on Trial; lossless on Startup and above (formats and sample rates not documented) | Not released | The vendor's own terms contradict each other |
| SOUNDRAW | not established | MP3 on all plans; WAV and stems on Artist Pro and above (bit depth and sample rate not documented) | Not released | The vendor's own terms contradict each other |
All 20
-
ACE-Step
An open-source music generation foundation model combining diffusion with a lightweight linear transformer, released May 2025 by ACE Studio and StepFun. It generates up to about four minutes
-
AIVA
Generates instrumental music from style presets and influence uploads for video, game and content creators, exporting MP3 and MIDI on lower tiers and WAV plus all formats on Pro. Its disting
-
Beatoven.ai maestro
Generates royalty-free background music and text-prompted sound effects for video, podcast, game and advertising work, downloadable as MP3 or WAV with a licence emailed on every download. Th
-
Eleven Music (music_v2)
Generates songs with vocals or instrumentals from a natural-language prompt, from user lyrics, or from a structured JSON 'composition plan' that specifies sections, timing and arrangement. S
-
ElevenLabs Sound Effects (eleven_text_to_sound_v2)
Generates sound effects, foley, ambiences and short musical elements from a text description, from 0.5 to 30 seconds per generation. A loop flag produces seamlessly looping output, and promp
-
Epidemic Sound AI tools (Studio, Search by video, Find similar)
Epidemic Sound is a licensed catalogue of human-made music and sound effects, not a generative music model. Its AI features are retrieval and editing tools: Search by video suggests tracks f
-
Google Lyria 3 Clip
The short-form Lyria 3 variant, model ID lyria-3-clip-preview, which produces a fixed 30-second clip from a text prompt with optional lyrics and image references. Intended for loops, preview
-
Google Lyria 3 Pro
Generates full songs with vocals and timed lyrics from a text prompt, optional user-supplied lyrics and up to ten reference images, at up to about three minutes with structural sections (int
-
Google Lyria RealTime
An experimental streaming music model (models/lyria-realtime-exp) that generates continuous instrumental music over a bidirectional WebSocket while you steer it live with weighted text promp
-
Loudly (VEGA / MANTA)
Generates music from text using its VEGA 1-3 instrumental models and MANTA 1 vocal model, then offers an AI Remixer, AI Mastering, a Stem Splitter (vocals, drums, bass, other), Studio FX and
-
Meta MusicGen (AudioCraft)
MusicGen is Meta's open text-to-music model, distributed in the AudioCraft research library alongside AudioGen (text-to-sound), MAGNeT, MusicGen Style (text-and-style conditioning), JASCO (c
-
Mubert
Generates royalty-free AI music from text prompts, images or a taxonomy of 200+ moods and themes, sold as a consumer 'Render' subscription and as a developer API with per-generation and stre
-
SOUNDRAW
Generates royalty-free instrumental and vocal tracks and lets you edit them in the browser: adjust length and energy, toggle instruments, blend genres, and mute/solo or change intensity bar
-
Stable Audio 3.0 Large
The flagship closed-weights model in the Stable Audio 3.0 family, released 20 May 2026 and aimed at music platforms and high-volume production. It generates variable-length music from text w
-
Stable Audio 3.0 Medium
The open-weights mid-size member of the Stable Audio 3.0 family, released 20 May 2026 for full song composition with tracks up to 6:20. Weights are published on Hugging Face under the Stabil
-
Stable Audio 3.0 Small and Small SFX
Two open-weight on-device variants released 20 May 2026. Small runs full music composition on a mobile device or consumer laptop with tracks up to two minutes; Small SFX is tuned for sound-e
-
Stable Audio Open Small
A 0.5B-parameter latent diffusion transformer released 14 May 2025 with Arm, generating up to 11 seconds of 44.1 kHz stereo audio from English text prompts on-device. It is better at sound e
-
Suno Studio 2.0
A browser-based generative DAW layered on top of Suno's generation models. It supports MIDI input and built-in synths, industry-standard audio effects including sidechain compression and con
-
Suno v5.5
Generates complete songs with vocals and lyrics from a text prompt, style tags and optionally user-written lyrics, or instrumentals only. v5.5 adds Voices (record and reuse your own singing
-
Udio v1.5 (Allegro)
Generates songs with vocals from a text prompt and optional user lyrics, in 32-second or 130-second segments that can be extended, remixed, inpainted and arranged on a Sessions timeline. v1.