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

68 models in instructions ranked by LMArena score · updated Nov 11, 2026

#ModelScore
1GLM-5.3Z.ai1483
2GLM-5.3-FlashZ.ai1480
3MiMo-V2.5-ProXiaomi MiMo1476
4DeepSeek-V4.1-FlashDeepSeek1473
5GLM-5.2Z.ai1467
6GLM-5.1Z.ai1452
7Kimi-K2.6Moonshot AI1451
8MiMo-V2.5Xiaomi MiMo1434
9gemma-4-31B-itGoogle1433
10Qwen3.8-27BQwen1431
11Kimi-K2.5Moonshot AI1431
12GLM-5Z.ai1428
13MiniMax-M3MiniMax1428
14InklingThinking Machines1427
15Hy3Tencent1426
16Beam-1.0-501B-A23Breflection1424
17Qwen3.5-397B-A17BQwen1424
18gemma-4-26B-A4B-itGoogle1421
19NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16NVIDIA1418
20Mistral-Medium-3.5-128BMistral AI1416
Show all 68 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.