Qwen2
Qwen2 is an open-weights AI model from Alibaba with an index score of 106 (GPT-4 = 100), rank 189 of 217.
10690–122Index score#189 of 217 models
–Price per million tokensno public API price found
–Context windowtokens it reads at once
7 Jun 2024ReleasedReplaced by Qwen2.5
Where it sitson the frontier
GPT-4 Turbo reached this score 7 months earlier.
Test resultsby area · tick = best by any model
Reasoning & knowledgeevidence
Mathsevidence
Codingevidence
Agentsno evidence yet
Professional workno evidence yet
Novel problemsno evidence yet
Long tasksevidence
Third-party results only, at the model’s best standard setting. Hover a bar for who ran it.
What others sayrank on 8 leaderboards
The Model Indexthird-party tests, our method#189of 217 · 106
LMArena · Mathsmaths questions in chat#227of 346 · 1274Epoch Capabilities Indexoverall capability from many benchmarks#158of 231 · 122.3LMArena · Creative writingcreative writing#261of 360 · 1223LMArena · Textoverall chat quality, judged by people#264of 362 · 1262LMArena · Codingcoding questions in chat#261of 357 · 1297LMArena · Instruction followingfollowing detailed instructions#268of 362 · 1243LMArena · Long promptslong, complex requests#268of 340 · 1262Best attop three among current models
Replaced by Qwen2.5; it no longer leads any category among current models.
Facts and sources
Released7 Jun 2024
Announcementqwenlm.github.io ↗
Release entryQwen2
Scored asQwen2-72B
Output speed–
Hugging Facenot linked