AI Providers

43 labs and providers tracked across five independent leaderboards.

Claude Mythos Preview97.4

65 models⌀ 48.2

16 models⌀ 52.13 open weights

GPT-5.6 Sol80.8

109 models⌀ 41.62 open weights

Qwen3.8 Max79.8

90 models⌀ 36.553 open weights

Gemini 3.1 Pro Preview74.0

65 models⌀ 42.314 open weights

Grok 4.20 Beta174.0

31 models⌀ 50.0

Deepseek V4 Pro High Preview73.0

33 models⌀ 45.419 open weights

Glm 5.2 Max73.0

23 models⌀ 51.711 open weights

Ernie 5.0 Preview 120372.0

5 models⌀ 54.6

Longcat Flash Chat 2602 Exp72.0

2 models⌀ 71.0

Amazon Nova Experimental Chat 26 02 1071.0

25 models⌀ 35.5

Nvidia Nemotron 3 Ultra 550b A55b Nvfp471.0

17 models⌀ 26.31 open weights

Hunyuan Hy3 Preview71.0

13 models⌀ 59.4

Minimax M2.570.0

9 models⌀ 41.24 open weights

Mistral Medium 250870.0

34 models⌀ 29.415 open weights

Step 3.5 Flash70.0

6 models⌀ 54.5

Muse Spark 1.169.9

43 models⌀ 38.518 open weights

Command A 03 202568.0

12 models⌀ 39.9

Olmo 3.1 32b Instruct67.0

7 models⌀ 54.51 open weights

Ling Flash 2.067.0

5 models⌀ 34.0

Granite 4.1 8b65.0

7 models⌀ 43.92 open weights

2 models⌀ 43.5

Phi 3 Medium 4k Instruct60.0

10 models⌀ 46.62 open weights

Sarvam 105B50.0

3 models⌀ 49.73 open weights

MiMo V2.5 Pro48.0

4 models⌀ 36.64 open weights

1 model⌀ 44.4

Thinking Machines

Inkling Small41.7

2 models⌀ 36.7

4 models⌀ 17.7

Seed 2.0 Pro40.0

2 models⌀ 29.3

Agnes 2 5 Pro Alpha39.7

1 model⌀ 39.7

Ring 2 6 1t31.3

1 model⌀ 31.3

K2 Think V222.0

1 model⌀ 22.0

Apriel V1 5 15b Thinker21.6

2 models⌀ 21.2

North Mini Code20.2

1 model⌀ 20.2

Trinity Large Thinking18.4

1 model⌀ 18.4

1 model⌀ 12.4

2 models⌀ 4.9

Jamba 1 5 Large4.8

2 models⌀ 3.6

Apertus 70b Instruct2.0

2 models⌀ 1.5

Async Flash V1 50.0

3 models⌀ 0.0

Cogito V2 1 Reasoning0.0

1 model⌀ 0.01 open weights