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

32 models in non-english ranked by LMArena score · updated Nov 11, 2026

#ModelScore
1Kimi-K3Moonshot AI1465
2GLM-5.3-FlashZ.ai1462
3Kimi-K3.1-ProMoonshot AI1450
4DeepSeek-V4.1-FlashDeepSeek1450
5Kimi-K2.6Moonshot AI1446
6InklingThinking Machines1436
7Kimi-K2.5Moonshot AI1432
8gemma-4-31B-itGoogle1430
9Qwen3.5-397B-A17BQwen1430
10Qwen3.8-27BQwen1424
11gemma-4-26B-A4B-itGoogle1421
12MiniMax-M3MiniMax1414
13Qwen3-VL-235B-A22B-InstructQwen1405
14MiMo-V2.5Xiaomi MiMo1404
15Inkling-SmallThinking Machines1402
16Qwen3.5-122B-A10BQwen1400
17Mistral-Large-4-1T-A52BMistral AI1397
18Inkling-2-SmallThinking Machines1390
19Qwen3.5-27BQwen1389
20NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16NVIDIA1386
21Muse-Glimmer-30Bmeta-models1385
22Qwen3.5-35B-A3BQwen1378
23Qwen3-VL-235B-A22B-ThinkingQwen1377
24GLM-4.6VZ.ai1359
25gemma-3-27b-itGoogle1334
26gemma-3-12b-itGoogle1318
27GLM-4.5VZ.ai1303
28gemma-3n-E4B-itGoogle1285
29Molmo2-8BAi21274
30gemma-3-4b-itGoogle1273
31Llama-4-Maverick-17B-128E-InstructMeta Llama1269
32Llama-4-Scout-17B-16E-InstructMeta Llama1252
Show the top 20

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.