Music & AudioMachine Learning4 min reading time

MiniMax Releases MiniMax-Music3: An Open-Weights Music Model Generating Complete Five-Minute Songs From Lyrics and a Structured Caption

MarkTechPost
Read full post
MiniMax launched MiniMax-Music3, an open-weights text-to-music model that generates full five-minute songs from lyrics and detailed captions. It produces high-quality stereo WAV outputs and is deployable immediately with available weights and code. The model architecture uses a hybrid large language model and continuous synthesis, targeting industries like gaming, advertising, and music tech.

More in Music & Audio

Music & Audio4 min read

Suno trained its v6 AI music models with help from Warner and BMG

Covered by 5 sources
Music & Audio2 min read

Roland is getting into generative AI music with Melody Flip

Covered by 2 sources
Music & Audio5 min read

Google Brings Lyria 3.5 Music Generation to the Gemini App and API

Covered by 2 sources