Original explanatory diagram
Qwen2.5-Omni put multimodal streaming on a joined input-output path
Qwen released Qwen2.5-Omni on March 27, 2025, describing one model that processes text, images, audio and video while streaming text and speech.
Read analysis