Qwen4-Max enters a time-limited preview on Alibaba’s Token Plan

Qwen has opened a preview of Qwen4-Max, a model in its new Qwen4 line. The official Qwen account said Qwen4 is coming, with stronger coding and agent capabilities, and that Qwen4-Max is now in preview on Alibaba’s Token Plan, with early access also on QwenWork, Qoder, the Qwen app and QwenStudio.

“Qwen4 is coming. Deeper knowledge. Richer world understanding. Stronger coding and agentic capabilities.”

— Qwen, in an 11 November announcement

Qwen Cloud’s documentation lists the model ID as qwen4-max-preview and describes it as the next-generation flagship multimodal model in the Qwen family, combining text generation, image and video understanding, and deep reasoning. The docs say it’s designed for coding, office work, front-end design and complex agent tasks, and that it supports web search, tool calling and structured outputs. The API is compatible with OpenAI Chat Completions, Responses, and Anthropic Messages.

The preview is available through Token Plan only; pay-as-you-go access isn’t supported. Official docs date the preview from 11 to 25 November, say no changes will be made during that period, and say the model will be retired when the preview ends.

The listed context length is 1 million tokens. The maximum output is 393,216 tokens, shared by the model’s thinking and the final answer. Thinking is on by default, with a reasoning_effort setting of low, medium or xhigh, which is the default. Starting with this model, max_tokens on the OpenAI Chat Completions API is treated as equivalent to max_completion_tokens, so the same budget covers both thinking and the answer. Earlier Qwen models keep their previous behaviour.

To call the preview through Token Plan, the docs say you need a dedicated API key from a Token Plan Personal or Team subscription, the base URL https://token-plan.maas.qwencloudapi.com, and the model ID qwen4-max-preview. Token Plan personal subscriptions are listed from US$6 a month, with a new-customer event price from as low as US$3, subject to the amount shown at checkout. Those are plan prices, not a separate per-token rate for Qwen4-Max.

Qwen4 is coming.

Deeper knowledge. Richer world understanding. Stronger coding and agentic capabilities.

Built for real-world tasks — from complex coding and knowledge work to front-end builds and professional software workflows.

Qwen4-Max is now in preview on Alibaba’s Token Plan. You can also get early access to the new model on QwenWork, Qoder, Qwen APP, and QwenStudio.

Put it to the test on a real problem. Show us what you build. 👀

Token Plan: https://t.co/LOB2Jdt3Sk

— Qwen (@Alibaba_Qwen) November 11, 2026

Qwen4 is coming.
Deeper knowledge. Richer world understanding. Stronger coding and agentic capabilities.
Built for real-world tasks—from complex coding and knowledge work to front-end builds and professional software workflows.
Qwen4-Max is now in preview on Alibaba’s Token Plan.

— Alibaba Cloud (@alibaba_cloud) November 11, 2026

Qwen hasn’t published a parameter count, a public benchmark table or downloadable weights for Qwen4-Max with this preview. The documentation’s last material limit is the retirement date: after 25 November, qwen4-max-preview is scheduled to come off Token Plan.