MiniMax Music 3.0
Open-weight text-and-lyrics-to-music model producing complete five-minute songs.
About
MiniMax Music 3.0 launched August 13, 2026 as a text-and-lyrics-to-music model that composes, arranges, performs, and produces a finished track in one pass, up to five minutes long at 32 kHz 16-bit stereo. It uses an 8B global LLM for long-range structure, a 0.6B local model for frame-level acoustic detail, flow matching synthesis, and a Flow-VAE decoder.
The weights are genuinely open on Hugging Face with ComfyUI-native workflows, but the MiniMax Community License requires the name MiniMax-Music3 displayed prominently in commercial products, with written authorization needed above $20M yearly revenue. Unlike the H3 video model, the Music 3 license has no territorial exclusion.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 105; Trending score 78/100 from researcher assessment.