Humanity's Last Exam without tools

Tool-free variant of Humanity's Last Exam that isolates a model's raw frontier reasoning.

In index
Knowledge
Best in knowledge

Models53

Top result73.2%Claude Fable 5.2

Top-3 spread12.3 pts

Year—

Result and price

30 of 53 results

Results

53 results

#ModelResult

01Claude Fable 5.273.2%

02Claude Opus 5.564.4%

03Claude Fable 5.160.9%

04Claude Mythos 559.0%

05Claude Sonnet 5.556.9%

06Claude Opus 556.3%

07Muse Spark 1.152.2%

08Kimi K3.1 Pro51.1%

09Sakana Fugu-Ultra50.0%

10Claude Opus 4.849.8%

11Sakana Fugu47.2%

12Claude Opus 4.7 (Adaptive)46.9%

13ERNIE 5.5 Flash Preview46.4%

14Claude Haiku 5.545.9%

15Gemini 3.1 Pro45.4%

16Ornith-1.5-397B44.6%

17Inkling 2 Small44.1%

18Qwen3.8 Max43.6%

19Kimi K343.5%

20MiniMax M3.1 Flash43.4%

20Hy4 preview43.4%

22Claude Sonnet 543.2%

23GPT-5.5 Pro43.1%

24Muse Spark42.8%

25GPT-5.4 Pro42.7%

26GPT-5.541.4%

27GLM-5.240.5%

28Claude Opus 4.640.0%

29GPT-5.439.8%

30Beam 1.0 501B-A23B36.0%

30Beam 1.0 501B-A23B FP836.0%

32Qwen3.8-Flash-Next35.9%

33Beam 1.0 501B-A23B NVFP434.8%

34MiMo-V2.5-Pro34.0%

35Nemotron 3.5 Super VL 120B A12B31.8%

36Inkling-Small31.6%

36Grok 4.2031.6%

38Nemotron 3.5 Super VL 120B A12B NVFP431.4%

39Qwen3.8-27B30.8%

40Inkling30.0%

41Solar Open 228.8%

42Nemotron 3.5 Bolt 74B-A9B BF1628.7%

43GPT-5.4 mini28.2%

44Nemotron 3 Ultra26.7%

45Ornith-1.5-35B-A3B25.6%

46Nemotron 3.5 Bolt 74B-A9B25.3%

47GPT-5.4 nano24.3%

48Ornith-1.5-9B20.2%

49Gemma 4 31B19.5%

50Spark-X2.6-4B17.9%

51Nemotron 3.5 Lightning 30B A3B NVFP410.5%

52Gemma 4 26B A4B8.7%

53Gemma 4 12B5.2%