M-A-P (Multimodal Art Projection)<\/strong> is an open-source AI research community. According to its m-a-p organization on Hugging Face<\/a>, its members work on a wide range of research, including pre-training of foundation models, large-scale data collection and processing, and applications to coding, reasoning and music creation, and the community is open to interested researchers. It is a publisher run by a group of researchers rather than a company.<\/p>\n Among generative models you can use locally, music generation<\/strong> stands out.<\/p>\n The music models and the text models have different terms.<\/strong> The music-related models from September 2026, such as YuE2 and MERT-v2, are non-commercial CC BY-NC 4.0<\/strong>; read the license text before using generated songs commercially. Text models such as OProver are Apache-2.0. Some models have no license field. This summary is not legal advice.<\/p>\n Sources: the m-a-p organization card<\/a>; the model card of m-a-p\/YuE2-3B<\/a>; the model list of the m-a-p organization on Hugging Face (all as of 2026-09-26).<\/em><\/p>\n Local Model Watch has published 1 article(s) on M-A-P’s own models, 2 on third-party fine-tunes and quantized builds based on them, and 0 from its official blog. The model list below is collected by code from Hugging Face every day. Part of our publisher index<\/a>; model families link to their family pages<\/a>.<\/p>\nWhat It Releases<\/h2>\n
\n
Licenses<\/h2>\n
Running Them Locally<\/h2>\n
\n
Our Coverage and Data<\/h2>\n
At a Glance<\/h2>\n