Qwen4-Max-Preview Is Here — But Only Until November 25

Alibaba Cloud began a time-limited preview of Qwen4-Max on November 11, 2026, giving developers access to a new flagship multimodal model for reasoning, coding, and tool-using workflows. The model ID is qwen4-max-preview, and it is hosted in Alibaba's Model Studio.

What Qwen4-Max-Preview offers

The capabilities are broad: text generation, image and video understanding, reasoning, and tool calling, with structured outputs. What makes the launch unusual is not the capability list. It is the access conditions. This is a preview, not open model weights and not a permanent pay-as-you-go API.

The access conditions

The preview runs from 10:00 on November 11 to 24:00 on November 25, 2026, in UTC+8, with retirement after that interval. Access is through Token Plan only, with no pay-as-you-go option documented.

Three practical consequences

That has three practical consequences. First, there is no public per-token dollar rate to budget against in the usual way. Second, any benchmark result from this window is useful for exploration but should not be the sole foundation for a long-lived production dependency. Third, a fallback model must be in place before November 25, not after.

Alibaba also announced on November 11 that the release of its next-generation large model, Qwen4, is imminent. The preview is best understood as an early look at that generation, not as a finished product launch.

How to evaluate it in two weeks

Two weeks is enough for a disciplined evaluation. Start by isolating the model behind an adapter, so endpoint credentials, response formats, and errors are separated from application code. Then test the things that actually break in production: complex tool-call sequences, schema-constrained output, multilingual prompts, and multimodal document interpretation. For export teams, that last category is especially relevant — a model that can read a scanned invoice, a product photo, and a Chinese-language specification in the same workflow is worth more than one that only handles clean English text.

Track task success and total billed Token Plan consumption, not just headline accuracy. Finally, version your stored results so they remain interpretable after the preview disappears.

Sources: SXF Brief, November 12, 2026; Alibaba Cloud documentation, November 11, 2026; Kim K.J. Daily Briefing, November 12, 2026.

How this gets paid for Testing like this takes a week nobody bills for. If you want more of it, the sponsorship page is how it keeps happening — no strings, no influence on the scores, and the running total is public.

We publish the pricing model, minimum commitment and export-readiness notes for every tool in the library, and we flag it when a vendor changes one without changing the headline price.