Alibaba launches Qwen4-Max, its latest AI model, on the cloud
Launch news
Alibaba’s Qwen4-Max artificial intelligence model has launched its final version on the company’s cloud service as of Monday, 16th. The system processes text, images, and videos but responds in written format. The launch follows a public preview that took place on November 11.
The new model continues the lineage that already includes Qwen3.8-Max, whose preview was unveiled in July. Alibaba now recommends the final version to developers who were previously testing the model.
What Qwen4-Max can do
Alibaba positions Qwen4-Max for programming tasks, office work, interface creation, and use by AI agents. The documentation mentions web search and integration with external tools, as well as support for API call formats from OpenAI and Anthropic.
The context window is one million tokens, which determines the amount of information considered in an interaction without limiting the response size. Tokens are the units used to process content and calculate usage costs.
Pricing and availability regions
In the international mode operated from Singapore, the rate is $1.50 per million input tokens and $4.50 per million output tokens.
The model’s reasoning tokens are included in the output charge. Prices vary by region, and the cost of each task depends on the amount of content processed and generated.
The launch is available in the United States, Germany, Japan, Hong Kong, mainland China, and Singapore. In August, Alibaba Cloud announced its first cloud region in Brazil, with two data centers, but the existence of this infrastructure does not confirm the availability of Qwen4-Max in the country.
Frequently Asked Questions
Does Qwen4-Max generate videos?
No. The model interprets videos submitted as input but produces responses only in text.
How can developers access Qwen4-Max?
The final version is available on Alibaba Cloud Model Studio under the identifier qwen4-max. The price and access conditions vary by region.








