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

63 models in math ranked by LMArena score · updated Nov 11, 2026

#ModelScore
63Olmo-3.1-32B-ThinkAi21305
62Olmo-3-32B-ThinkAi21315
61gpt-oss-20bOpenAI1317
60MiniMax-M2MiniMax1353
59GLM-4.5VZ.ai1354
58Qwen3-30B-A3BQwen1355
57GLM-4.7-FlashZ.ai1355
56QwQ-32BQwen1359
55Ling-flash-2.0inclusionAI1365
54Trinity-Large-ThinkingArcee AI1365
53NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA1366
52NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA1370
51INTELLECT-3Prime Intellect1376
50MiniMax-M2.5MiniMax1377
49NVIDIA-Nemotron-3-Super-120B-A12B-BF16NVIDIA1379
48gpt-oss-120bOpenAI1389
47DeepSeek-R1DeepSeek1393
46MiMo-V2-FlashXiaomi MiMo1395
45GLM-4.5-AirZ.ai1396
44Qwen3-235B-A22BQwen1397
Show all 63 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.