Qwen3-Coder
Qwen3-Coder is an open-weights AI model from Alibaba, released 22 Jul 2025, with no index score yet.
–Index scoreNo third-party results yet
$0.30/ $1.0Price per million tokensinput / output · via OpenRouter
262KContext windowtokens it reads at once
22 Jul 2025Released1 year 4 months ago
Where it sitson the frontier
Not plotted: no third-party score yet.
Test resultsby area · tick = best by any model
Reasoning & knowledgeno evidence yet
Mathsno evidence yet
Codingevidence
Agentsno evidence yet
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 7 leaderboards
LMArena · Codingcoding questions in chat#126of 357 · 1457LMArena · Creative writingcreative writing#130of 360 · 1362LMArena · Long promptslong, complex requests#130of 340 · 1408LMArena · Instruction followingfollowing detailed instructions#139of 362 · 1383LMArena · Textoverall chat quality, judged by people#148of 362 · 1387LMArena · Mathsmaths questions in chat#153of 346 · 1375LMArena · WebDevbuilding web apps and front ends#115of 133 · 1274
Best 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 2 · per 1M tokens, in / out
Run it yourself
Official weightsQwen/Qwen3-Coder-480B-A35B-Instruct↗Host prices from OpenRouter; going direct can differ.
Facts and sources
Released22 Jul 2025
Announcementqwenlm.github.io ↗
Release entryQwen3-Coder
Scored asQwen3-Coder-480B-A35B
Output speed–
Hugging FaceQwen/Qwen3-Coder-480B-A35B-Instruct ↗