Qwen4-Max-Preview opens on Token Plan

Alibaba Cloud Model Studio's page for qwen4-max-preview, last updated 11 November 2026, calls Qwen4-Max-Preview the next-generation flagship multimodal model in the Qwen family. It combines text generation, image and video understanding, and deep reasoning, for coding, office productivity, frontend design, and complex agent tasks. The page says it supports web search, tool calling, and structured outputs, and that it works with the OpenAI Chat Completions API, the Responses API, and the Anthropic Messages API.

The preview is Token Plan only. Pay-as-you-go is not supported. Alibaba says the window runs from 10:00 on 11 November 2026 to 24:00 on 25 November 2026 (UTC+8), that the model will not change during that window, and that it will be retired when the preview ends. Context length is 1 million tokens. Maximum output is 393,216 tokens, shared by thinking and the final answer. Thinking is on by default. reasoning_effort can be low, medium, or xhigh, and xhigh is the default. The page says to leave thinking_budget unset when reasoning_effort is set.

On this model, max_tokens means the same thing as max_completion_tokens. Both cap thinking and the answer together, so a limit sized only for the answer can cut the reply short. Calls use a Token Plan key and the base URL https://token-plan.ap-southeast-1.maas.aliyuncs.com, with the model id qwen4-max-preview.

Who it's for: a developer on Token Plan who wants a preview of Qwen's flagship for a coding or agent task before 25 November 2026.

Fits: Best for Vibe Coding stays Cursor. Best for Writing stays Claude. A preview API model leaves both jobs where they are. The coding guide is unchanged. The verdict is unchanged.