Saturday, November 21, 2026
Translate
PrivacySubmit newsShare your story Newsletter Advertise
Breaking
AI news

Qwen4 Max Processes Entire Repositories in a Single Call on Together AI

Alibaba's flagship 2.4-trillion-parameter model is now available on Together Serverless Inference.

Qwen4 Max Processes Entire Repositories in a Single Call on Together AI
Image: Together AI

Alibaba's flagship Qwen4 Max model is now live on Together Serverless Inference, offering a 1-million-token context window and automatic prompt caching.

Availability and Specifications

Qwen4 Max is a sparse mixture-of-experts model with 2.4 trillion parameters, of which 71 billion are active for each token. Alibaba released Qwen4 Max on November 16, 2026, and describes it as the most capable model in the Qwen family to date.

The model is available on Together Serverless Inference under the model string Qwen/Qwen4-Max. Requests must set stream=True because the model accepts streaming requests only, and thinking is enabled by default with thinking tokens counted as output tokens.

The TechUpscale Brief. The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

Key facts and where they come from
  • Qwen4 Max is available on Together AI with a 1M-token context window.
    Alibaba's flagship Qwen model is available on Together Serverless Inference with a 1M-token context window.
  • Alibaba released Qwen4 Max on November 16, 2026.
    Alibaba released Qwen4 Max on November 16, 2026, and describes it as the most capable model in the Qwen family to date.

The catch

  • Requests to Qwen4 Max on Together AI must set streaming to true.
  • Thinking tokens are enabled by default and count as output tokens.

The open question: How will developers adapt their workflows when entire repositories can be processed in a single inference call?

Tell us what you think · Discuss it in the AI room

Reader poll: Have you tested models with a 1M-token context window on entire repositories?

A show of hands from readers, not a scientific poll. Say why in the comments.

Add TechUpscale as a preferred source on Google

The TechUpscale Brief

The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

I'm most interested in

How will developers adapt their workflows when entire repositories can be processed in a single inference call?

Comments are moderated. Be civil and stay on the story: no personal attacks, hate, spam or ads, and no one else's private details. Your email is never shown. Privacy policy.

More AI