Benchgen

Humanity's Last Exam — Results

RankModelScore
1claude-sonnet-557.4
2ornith-1-5-397b56.1
3kimi-k356
4seed-2-1-pro55.7
5glm-5-3-flash55.3
6glm-5-254.7
7seed-2-1-turbo54.6
8kimi-k2-thinking-090551
9fugu-ultra50
10qwen3-5-27b48.5
11deepseek-v4-pro-max48.2
12qwen3-5-122b-a10b47.5
13qwen3-5-35b-a3b47.4
14fugu47.2
15gemini-3-1-pro46.44
16inkling46
17deepseek-v4-flash-max45.1
18gpt-5-444.32
19qwen3-8-max43.6
20gemini-3-flash43.5
21claude-sonnet-543.2
22glm-4-742.8
23deepseek-v4-pro42.7
24gpt-5-541.4
25qwen3-7-max41.4

Humanity's Last Exam — Leaderboard

1 phaseActive

Humanity's Last Exam Leaderboard

No self-reported entries.