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 ranked by LMArena score · updated Nov 11, 2026

#ModelScore
68Olmo-3.1-32B-ThinkAi21271
67gpt-oss-20bOpenAI1288
66granite-4.2-3bIBM Granite1292
65Olmo-3-32B-ThinkAi21299
64Qwen3-30B-A3BQwen1317
63granite-4.2-8bIBM Granite1319
62QwQ-32BQwen1329
61NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA1329
60GLM-4.5VZ.ai1333
59Trinity-Large-ThinkingArcee AI1340
58Qwen3-32BQwen1340
57MiniMax-M2MiniMax1340
56NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA1348
55GLM-4.7-FlashZ.ai1351
54INTELLECT-3Prime Intellect1356
53MiniMax-M2.5MiniMax1359
52Ling-flash-2.0inclusionAI1361
51granite-4.2-30bIBM Granite1362
50gpt-oss-120bOpenAI1366
49Qwen3-Next-80B-A3B-ThinkingQwen1368
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.