Through an API

Open and closed models answer the same suite-standard 1.0 through their APIs, ranked together. 1 ranked model.

#ModelPass rate
1DeepSeek V4.1 Flash 1017Open weights100%

Model leaderboard

86 models in multi-turn ranked by LMArena score · updated Nov 11, 2026

#ModelScore
1MiMo-V2.5-ProXiaomi MiMo1476
2GLM-5.3Z.ai1474
3GLM-5.1Z.ai1472
4GLM-5.3-FlashZ.ai1471
5GLM-5.2Z.ai1469
6DeepSeek-V4.1-FlashDeepSeek1462
7GLM-5Z.ai1456
8Kimi-K2.6Moonshot AI1454
9gemma-4-31B-itGoogle1452
10GLM-4.7Z.ai1446
11MiMo-V2.5Xiaomi MiMo1445
12Qwen3.5-397B-A17BQwen1445
13Kimi-K2.5Moonshot AI1444
14InklingThinking Machines1442
15gemma-4-26B-A4B-itGoogle1442
16MiniMax-M3MiniMax1439
17Hy3Tencent1438
18Qwen3.8-27BQwen1436
19Qwen3-235B-A22B-Instruct-2507Qwen1432
20Mistral-Large-3-675B-Instruct-2512Mistral AI1429
Show all 86 models

Score by size

Plot against
Best for its size

Scores: LMArena, CC BY 4.0. Measured on served versions, not local builds. A model with several served variants shows its best.