Alibaba Cloud opened Qwen4-Max Preview on November 11 for its Model Studio Token Plan Personal and Team subscribers. Developers use a dedicated plan key and Beijing endpoint. Pay-as-you-go access is unavailable. Alibaba says the preview model will be retired when the preview ends at 24:00 on November 25 in Beijing (16:00 UTC).
The new flagship accepts text, images and video and generates text. It has a one-million-token context window and supports tool calls and structured output. The maximum output is 393,216 tokens, shared between thinking and the final answer. In Chat Completions, either output-limit parameter now caps both, so a setting previously sized for the answer alone can cut it short. Alibaba says the change does not affect existing Qwen models.
How the preview uses credits
Model calls draw from a Token Plan credit balance. Alibaba says preview usage can cost as little as 20% of the normal credit charge, with the two discounts bringing usage to as little as 8% of the standard credit charge between 10 p.m. and 8 a.m. Beijing time. The promotion lasts through the preview. Standard subscriptions start at 39 yuan a month for Personal and 150 yuan per seat a month for Team. The preview documentation gives no date for a successor or a standard per-token price.