Qwen3.5
Qwen3.5 is an open-weights AI model from Alibaba with an index score of 144 (GPT-4 = 100), rank 85 of 217.
144127–161Index score#85 of 217 models
$0.45/ $3.0Price per million tokensinput / output · via OpenRouter
262KContext windowtokens it reads at once
16 Feb 2026Released9 months ago
Where it sitson the frontier
o3 reached this score 10 months earlier.
Test resultsby area · tick = best by any model
Reasoning & knowledgeevidence
Mathsevidence
Codingno evidence yet
Agentsevidence
Professional workevidence
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 10 leaderboards
The Model Indexthird-party tests, our method#85of 217 · 144
LMArena · Instruction followingfollowing detailed instructions#37of 362 · 1468LMArena · Long promptslong, complex requests#37of 340 · 1482LMArena · Creative writingcreative writing#42of 360 · 1446LMArena · Codingcoding questions in chat#48of 357 · 1513LMArena · Textoverall chat quality, judged by people#50of 362 · 1465LMArena · Mathsmaths questions in chat#52of 346 · 1468Epoch Capabilities Indexoverall capability from many benchmarks#76of 231 · 147.9LMArena · Visionunderstanding images#51of 145 · 1246LMArena · WebDevbuilding web apps and front ends#89of 133 · 1400Best attop three among current models
Not in the top three among current models on any category we track.
Where to use itprices checked 18 Nov 2026
API providers 8 · per 1M tokens, in / out
Run it yourself
Official weightsQwen/Qwen3.5-397B-A17B↗Host prices from OpenRouter; going direct can differ.
Facts and sources
Released16 Feb 2026
Announcementqwenlm.github.io ↗
Release entryQwen3.5 (397B open + 3.5-Plus)
Scored asQwen3.5 397B-A17B
Output speed–
Hugging FaceQwen/Qwen3.5-397B-A17B ↗