Qwen3.8-Max is here: Alibaba's 2.4-trillion-parameter model codes on its own for 10 days
Alibaba has released Qwen3.8-Max, a 2.4-trillion-parameter mixture-of-experts model (95 billion active per token), two weeks after its preview. Company benchmarks show it beating GPT-5.6 Sol on instruction following and research-style tasks while still trailing Claude Fable 5 on hard software engineering; the API costs $2 per million input tokens and $6 per million output, and open weights are promised within about a week. Alibaba's showcase: the model ran a software project autonomously for 10 days, opening its own issues, running tests and merging its own pull requests — and Alibaba shares jumped about 6% on the launch.
Why it mattersThe frontier race is no longer American-only: a Chinese flagship now matches or beats Western leaders on several agent benchmarks at a fraction of the price — and its weights are about to become downloadable. For developers, that means real pressure on API prices; for everyone else, models that work unattended for days are quietly becoming normal.
▶ Related video: QWEN 3.8 Max Is Insane! SECOND To Fable? New Open Model King? (Fully Tested)
Read in the app — free, in 9 languages
Related stories
Rumour: the anonymous model that just topped a coding benchmark, for free, is said to be Zhipu's unreleased flagship2026-08-21DeepSeek's cheap workhorse can now see — and on agent tasks that need eyes it says it is close to Anthropic's best
2026-08-21Nvidia is paying $6 billion for the machine that builds a rival's models — and hiring 109 of the people who ran it
2026-08-21Given four hours and a GPU to improve the way AI is trained, the best agent scored 0.25 out of 1 — and most never tried
2026-08-21ChatGPT can now read your iMessages and send them — and the setting that lets it skip asking is the one OpenAI warns about
2026-08-21