Meta Force Space
BTC $64,946.00 -0.50% ETH $1,816.75 -0.97% SOL $71.12 -1.99% XRP $0.9513 -2.41% BNB $590.07 -0.96% DOGE $0.0587 -1.90%
← Back to the news

Alibaba's Qwen 4 Max Leads China's AI Agent Race—But You Can't Download It

In brief

  • Alibaba has released Qwen 4 Max, a 2.4-trillion-parameter model with a one-million-token context window and native visual understanding.
  • It leads Chinese rivals on Arena's agent leaderboard and ranks sixth globally for AI-built websites, with particular gains in visual design.
  • Qwen 4 Max costs $1.50 per million input tokens and $4.50 per million output tokens, but its weights are closed. Smaller open-weight Qwen 4 models are promised.

Alibaba released Qwen 4 Max on Monday, its most powerful AI model yet and the first flagship built on its new Qwen 4 architecture.

It is also the highest-ranked Chinese model on Arena's agent leaderboard, putting the company ahead of Tencent, Moonshot AI and DeepSeek in that particular test.

The new model has 2.4 trillion parameters, can process text, images and video, and accepts up to one million tokens of context. Parameters are all the different dials a model can handle and tweak when processing information. Tokens are basically the smallest unit of information a model can handle. Its current configuration is enough room for lengthy codebases or stacks of documents.

$625.67K volume
Closes: Jan 1, 2027, 04:59 UTC
Closes: Jan 1, 2027, 04:59 UTC

Alibaba uses a mixture-of-experts design, which keeps most of those parameters idle for any given word it processes. Qwen 4 Max activates 71 billion at a time, compared with 95 billion for Qwen 3.8 Max, allowing a very large model to use less computing power per token.

It can code, and it has an eye for design

On Agent Arena, Qwen 4 Max placed 14th among individual models and made Alibaba the fifth-ranked laboratory overall.

The leaderboard evaluates real agent sessions using task completion, user satisfaction and how well models respond to corrections, then scores each model by its net percentage improvement over an average competitor. Qwen 4 Max posted a preliminary 7.46% score across 14,727 sessions, the highest among Chinese entries.

Its visual work may be the more interesting part for regular users. The Code Arena web-development ranking compares websites built by AI models using head-to-head user votes and a points-based rating system. Qwen 4 Max debuted sixth globally at 1,750 points, up 74 from its predecessor, and placed eighth in categories for branding and reference-based design.

That puts it in the same neighborhood as the best American models at turning a request, a screenshot or a rough idea into a working interface that actually looks good.

Alibaba's demos go further, showing the model building game environments and translating factory layouts into 3D models, although those are company-selected examples rather than independent tests.

The visual capability is for understanding images and video and producing code or other text-based outputs. It is not the same as releasing a new image generator.

Cheaper intelligence, with a catch

Qwen 4 Max costs $1.50 per million input tokens and $4.50 per million generated tokens through QwenCloud, a 25% reduction from the previous flagship's $2 and $6 rates. Reused input served from an automatic cache costs $0.075 per million tokens, which can make repetitive workflows substantially cheaper.

Alibaba says its internal testing across hundreds of professional specialties found that the model used 35% fewer tokens and reduced average task costs by roughly 60% compared with Qwen 3.8 Max. In other words, it should be cheaper to execute tasks on this model even against competitors with a similar price..

The bigger limitation for the community is that Qwen 4 Max is closed-weight. Developers can use it through an API or Alibaba's apps, but they cannot download its internal parameters, modify it or run it privately on their own hardware.

Qwen has become a foundation for the open-model ecosystem. An August report from Hugging Face counted 151,448 Qwen-based derivative models on its platform, about 4.7 times the number built specifically on Meta's Llama family. Those include community fine-tunes and optimized versions that power local assistants and specialized tools.

Alibaba opened the weights of Qwen 3.8 Max earlier this year, despite having kept some previous Max models proprietary. It also previewed elements of the Qwen 4 architecture with smaller models months before Monday's release.

The launch lands in a crowded Chinese race. Moonshot's Kimi K3 pushed enormous open-weight models toward Western performance, while DeepSeek has been competing hard on API prices.

Qwen 4 Max leads some Chinese rankings, but American labs still hold the top spots in Arena's overall agent and web-development tables. They are also considerably more expensive.

Alibaba says smaller Qwen 4 models will be released with open weights, giving developers a chance to build on this generation rather than simply rent its biggest model. It has not announced their sizes or release dates.

Daily Debrief Newsletter

Start every day with the top news stories right now, plus original features, a podcast, videos and more.

Join 72,000+ readers
Create an account to save your articles.
Add on Google
Add Decrypt as your preferred source to see more of our stories on Google.

Originally published by Decrypt on

Read the original on Decrypt ↗

Text and images are the property of Decrypt and are reproduced here with attribution and a link to the original publication.

More stories

All the latest news