Qwen/Qwen2.5-Omni-3B
Any-to-Any
•
6B
•
Updated
•
256k
•
302
Models that can flexibly convert across modalities (text, image, audio, video). Ideal for researchers exploring unified multimodal-AI.