AI Guides
PowerfulAI.tools Editorial Team
0 Comments
By PowerfulAI.tools Editorial Team
![]()
Alibaba has released a preview of Qwen4-Max, which it describes as the next-generation flagship model in its Qwen family of AI models.
The preview opened on 11 November 2026 and is deliberately short. It runs only until 25 November 2026, and access is limited to a paid coding subscription. Here is what launched, who can use it, what it costs and what it means if you are not a developer.
According to Alibaba Cloud's Model Studio documentation, Qwen4-Max-Preview is a multimodal model. That means it works with more than just text. Alibaba lists these main capabilities:
Alibaba says the model is designed for coding, office work, front-end (website and app interface) design and complex agent tasks, where the AI carries out several steps on its own.
The model has a context window of 1 million tokens, which is roughly the amount of text it can consider at once. Its maximum output is 393,216 tokens, and that limit is shared between its thinking and its final answer.
Alibaba's documentation does not publish any benchmark results or comparisons with other AI models, so we cannot tell you how it stacks up against ChatGPT, Gemini or Claude. It also does not say whether the model's weights will be released for people to download.
Qwen4-Max-Preview is not a general release. Alibaba's documentation says it is "available through Token Plan only", and pay-as-you-go API access is not supported.
Token Plan is Alibaba Cloud's subscription for AI coding and agent tools. Alibaba lists Claude Code, Cursor, Qwen Code, Qoder and OpenClaw as supported tools. The plan is meant for interactive use inside those tools. Alibaba does not allow it to be used for automation scripts, app backends or batch jobs.
Availability by country is not clear-cut. The international documentation for the model gives a Singapore-region connection address. However, Alibaba's overview page for Token Plan Personal Edition says that edition is currently available only in the China (Beijing) region. If you are in Australia and want to try it, check the sign-up page carefully before paying.
The preview window is fixed. Alibaba says it runs from 10:00 on 11 November to midnight on 25 November 2026 (China time, UTC+8), that no changes will be made during that period, and that the preview model will be retired when it ends.
Alibaba's Token Plan Personal Edition page lists these standard monthly prices in Chinese yuan:
Alibaba is also running limited-time offers. New subscribers to Lite and Essential can get a lower first price until 30 November 2026, and during the preview Qwen4-Max-Preview uses fewer plan credits than the standard rate. These are temporary promotions, not normal pricing. Alibaba has not published a price for any final Qwen4-Max release.
For most people, nothing changes yet. Qwen4-Max-Preview is a short trial aimed at developers who use AI coding tools, and it sits behind a paid plan with unclear availability outside China.
The bigger signal is that Alibaba's Qwen4 generation is now on the way. Qwen models already power the free Qwen chat assistant, and Alibaba has released some earlier Qwen models for anyone to download. If a full Qwen4-Max release follows this preview, it could add another option alongside the big US-made assistants. Until Alibaba publishes benchmarks and wider availability, it is too early to say how it compares.
If you are a developer who already uses Claude Code, Cursor or Qwen Code, the two-week window is a low-commitment way to try the model. Everyone else can wait for the full release.
We checked this article on 12 November 2026 against Alibaba Cloud's official Model Studio documentation for Qwen4-Max-Preview (https://www.alibabacloud.com/help/en/model-studio/qwen4-max-preview) and its Token Plan Personal Edition overview (https://help.aliyun.com/en/model-studio/token-plan-personal-overview). Claims we could not confirm on Alibaba's own pages, including reports of the model appearing in the Qwen chat app, have been left out. Logo: Qwen.
Leave a Comment