Alibaba ships Qwen3.7 Flash with 1M context and video input
Alibaba's new vision model reads video across a million token window at $0.03 per million input tokens, though the headline rate only holds under 32K.
This is some text inside of a div block.