Alibaba's new top Qwen model: clear gains on terminal coding work and hard reasoning tests over 3.6-Plus, plus a much larger memory — at the same price.
Compared with Qwen 3.6-Plus · the model it replaces
Solves 7 in 10 hands-on coding tasks in a terminal
Terminal-Bench 2.0 — doing real work in a command-line terminal
69.7%
▲ up from 61.6% · +8 points
Answers about 4 in 10 of the hardest expert exam questions
Humanity's Last Exam — graduate-level questions across many fields
38.1%
▲ up from 28.9% · +9 points
Costs the same to run as before
Price per 1M tokens — what developers pay, input / output
$2.50 / $7.50
same price as Qwen 3.6-Plus
Qwen 3.8-Max is the current model in this line. The card shows how long it has been since the last release against the line's usual gap and its recent ones — the rhythm, not a date.
Newest first. See the full line with every card →
One email a week: what AI models shipped, the actual differences vs the previous version, and what's coming next. No filler.