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

67 models in expert ranked by LMArena score · updated Nov 11, 2026

#ModelScore
67gpt-oss-20bOpenAI1259
66Olmo-3-32B-ThinkAi21273
65Olmo-3.1-32B-ThinkAi21294
64granite-4.2-3bIBM Granite1305
63Qwen3-30B-A3BQwen1314
62INTELLECT-3Prime Intellect1323
61QwQ-32BQwen1323
60DeepSeek-R1DeepSeek1337
59MiniMax-M2MiniMax1340
58GLM-4.5VZ.ai1351
57Ling-flash-2.0inclusionAI1353
56NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA1355
55gpt-oss-120bOpenAI1355
54GLM-4.7-FlashZ.ai1358
53Trinity-Large-ThinkingArcee AI1359
52NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA1360
51Qwen3-32BQwen1362
50Qwen3-235B-A22BQwen1369
49GLM-4.5-AirZ.ai1370
48GLM-4.6VZ.ai1372
Show all 67 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.