Model
151d agoQwen3.5‑Omni Writes Code From Video
Alibaba’s new Qwen3.5‑Omni model demonstrates unprecedented multimodal learning, generating code from spoken instructions and video input without explicit training. By outperforming rivals on audio tasks, it showcases the potential of integrated vision, speech, and language capabilities.
The Signal
This breakthrough signals a shift toward more autonomous, cross‑modal AI systems that can adapt to diverse media worldwide.