Ming-Omni

Ant Group's native multimodal model built on the Ling backbone. Handles vision, speech, audio, and music.