Qwen4-Max

Alibaba released Qwen4-Max, its new flagship model, through its QwenCloud API and apps, but without downloadable open weights.

Qwen4-Max is a mixture-of-experts model with 2.4 trillion total parameters (about 71 billion active per token), native image and video understanding and a one-million-token context window. Alibaba says it beats Qwen3.8-Max on its own Work360 benchmark while cutting the average cost per task by about 60%, and it claims near-top coding scores; those are company figures, and TNW notes it trails some rivals on Terminal-Bench. API pricing is $1.50 per million input and $4.50 per million output tokens. It is available on QwenCloud and in QwenWork, QwenStudio and Qoder. Unlike many earlier Qwen releases, the weights are closed; Alibaba says smaller Qwen4 models will get open weights, with no date. Alibaba also highlights new safety training, including removing offensive-security training data.

Status
Released
Date
Nov 16, 2026
Maker
Alibaba
Type
AI model
Price
$1.50 / $4.50 per million input / output tokens (QwenCloud API)
Categories
AI

Sources

Launch dates can change. Check the sources for the latest. Last reviewed Nov 17, 2026.

Related rumors