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

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

#ModelScore
1GLM-5.3Z.ai1462
2GLM-5.3-FlashZ.ai1462
3GLM-5.2Z.ai1458
4DeepSeek-V4.1-FlashDeepSeek1450
5GLM-5.1Z.ai1448
6MiMo-V2.5-ProXiaomi MiMo1448
7Kimi-K2.6Moonshot AI1446
8InklingThinking Machines1436
9Kimi-K2.5Moonshot AI1432
10GLM-5Z.ai1431
11gemma-4-31B-itGoogle1430
12Qwen3.5-397B-A17BQwen1430
13NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16NVIDIA1428
14Hy3Tencent1426
15GLM-4.6Z.ai1426
16Qwen3.8-27BQwen1424
17gemma-4-26B-A4B-itGoogle1421
18GLM-4.7Z.ai1417
19GLM-4.5Z.ai1417
20Mistral-Large-3-675B-Instruct-2512Mistral AI1414
21MiniMax-M3MiniMax1414
22Beam-1.0-501B-A23Breflection1412
23Qwen3-235B-A22B-Instruct-2507Qwen1409
24Qwen3-Next-80B-A3B-InstructQwen1407
25Mistral-Medium-3.5-128BMistral AI1405
26Qwen3-VL-235B-A22B-InstructQwen1405
27MiMo-V2.5Xiaomi MiMo1404
28LongCat-Flash-ChatMeituan LongCat1404
29Inkling-SmallThinking Machines1402
30Qwen3.5-122B-A10BQwen1400
31Qwen3-235B-A22B-Thinking-2507Qwen1397
32Mistral-Large-4-1T-A52BMistral AI1397
33MiMo-V2-FlashXiaomi MiMo1392
34Inkling-2-SmallThinking Machines1390
35Hy3-previewTencent1390
36Qwen3.5-27BQwen1389
37NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16NVIDIA1386
38Step-3.5-FlashStepFun1385
39Muse-Glimmer-30Bmeta-models1385
40Qwen3-235B-A22BQwen1384
41MiniMax-M2.7MiniMax1383
42Qwen3.5-35B-A3BQwen1378
43Qwen3-VL-235B-A22B-ThinkingQwen1377
44Qwen3-30B-A3B-Instruct-2507Qwen1372
45GLM-4.5-AirZ.ai1366
46GLM-4.6VZ.ai1359
47NVIDIA-Nemotron-3-Super-120B-A12B-BF16NVIDIA1355
48gpt-oss-120bOpenAI1350
49Qwen3-Coder-480B-A35B-InstructQwen1345
50Qwen3-Next-80B-A3B-ThinkingQwen1343
51granite-4.2-30bIBM Granite1340
52MiniMax-M2.5MiniMax1338
53Ling-flash-2.0inclusionAI1335
54GLM-4.7-FlashZ.ai1330
55INTELLECT-3Prime Intellect1328
56Trinity-Large-ThinkingArcee AI1326
57NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA1320
58Trinity-Large-PreviewArcee AI1318
59Qwen3-32BQwen1317
60MiniMax-M2MiniMax1313
61QwQ-32BQwen1305
62GLM-4.5VZ.ai1303
63granite-4.2-8bIBM Granite1300
64NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA1297
65Qwen3-30B-A3BQwen1295
66Olmo-3.1-32B-InstructAi21276
67Llama-4-Maverick-17B-128E-InstructMeta Llama1269
68gpt-oss-20bOpenAI1268
69granite-4.2-3bIBM Granite1268
70granite-4.1-8bIBM Granite1261
71Llama-4-Scout-17B-16E-InstructMeta Llama1252
72Qwen2.5-72B-InstructQwen1252
73Llama-3.1-405B-InstructMeta Llama1248
74Llama-3.3-70B-InstructMeta Llama1236
75Llama-3.1-70B-InstructMeta Llama1219
76granite-4.0-h-smallIBM Granite1217
77Qwen2.5-Coder-32B-InstructQwen1205
78Llama-3.1-8B-InstructMeta Llama1148
79QwQ-32B-PreviewQwen1130
80Mixtral-8x22B-Instruct-v0.1Mistral AI1128
81granite-3.1-8b-instructIBM Granite1099
82granite-3.1-2b-instructIBM Granite1068
83granite-3.0-8b-instructIBM Granite1037
84granite-3.0-2b-instructIBM Granite1033
85Llama-3.2-3B-InstructMeta Llama1019
86Llama-3.2-1B-InstructMeta Llama973
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.