Qwen3.8-Max
Qwen3.8-Max is a closed AI model from Alibaba with an index score of 167 (GPT-4 = 100), rank 32 of 217.
167151–183Index score#32 of 217 models
–Price per million tokensno public API price found
–Context windowtokens it reads at once
3 Aug 2026Released4 months ago
Where it sitson the frontier
GPT-5.5 reached this score 3 months earlier.
Test resultsby area · tick = best by any model
Reasoning & knowledgeevidence
Mathsevidence
Codingevidence
Agentsevidence
Professional workno evidence yet
Novel problemsno evidence yet
Long tasksno evidence yet
Third-party results only, at the model’s best standard setting. Hover a bar for who ran it.
What others sayrank on 11 leaderboards
The Model Indexthird-party tests, our method#32of 217 · 167
LMArena · Visionunderstanding images#3of 145 · 1299LMArena · Searchanswering with web search#1of 41 · 1260LMArena · Mathsmaths questions in chat#9of 346 · 1505LMArena · Creative writingcreative writing#15of 360 · 1472LMArena · Long promptslong, complex requests#18of 340 · 1496LMArena · Textoverall chat quality, judged by people#22of 362 · 1484LMArena · Instruction followingfollowing detailed instructions#24of 362 · 1478LMArena · Codingcoding questions in chat#27of 357 · 1523LMArena · WebDevbuilding web apps and front ends#15of 133 · 1676Epoch Capabilities Indexoverall capability from many benchmarks#30of 231 · 157.6Best attop three among current models
Where to use itprices checked 18 Nov 2026
Host prices from OpenRouter; going direct can differ.
Facts and sources
Released3 Aug 2026
Announcementqwenlm.github.io ↗
Release entryQwen3.8-Max (preview 07-19; weights 08-12)
Scored asQwen 3.8 Max
Output speed–
Hugging Faceclosed weights