Alibaba makes Qwen4-Max generally available
The first model of the Qwen4 family is now offered through Alibaba’s API, with a context of 1 million tokens, reasoning, image and video input and built-in tools.
In the Daily Brief of 16 November 2026
Filed under Large language models, AI industry
Alibaba’s documentation describes Qwen4-Max as a model for coding, professional work, long-running agents and scientific tasks. It accepts text, images, video and PDFs, supports function calling, web search, a code interpreter and the Model Context Protocol, and can return up to 262,144 tokens of reasoning and answer combined.
The model is served from Singapore, Beijing and several global regions. Fine-tuning and private deployment are not supported. Performance figures come from Alibaba’s own launch material and have not been checked independently.