TokenCostEstimate bill →
← All benchmarks·Source: Epoch AI ↗

Humanity's Last Exam leaderboard

Current leader: Gemini 3.1 Pro at 43.74 (% accuracy).

Cost vs quality

Each dot is a model. The dashed line is the value frontier — top-left is the best trade-off (high score, low cost).

Closed model Open-weight model

Full leaderboard

#ModelScore% accuracyOutput $/1MReleased
1
Gemini 3.1 Pro
43.74$12.00Feb 2026
2
GPT-5.4 Pro
41.51$180.00Mar 2026
3
Muse Spark
37.56—Apr 2026
4
Gemini 3 Pro
34.37—Nov 2025
5
GPT-5.4
33.03$15.00Mar 2026
6
Claude Opus 4.7
32.98$150.00Apr 2026
7
Claude Opus 4.6
31.13$25.00Feb 2026
8
GPT-5 Pro
28.19—Oct 2025
9
GPT-5.2
24.16—Dec 2025
10
GPT-5
21.55—Aug 2025
11
Claude Opus 4.5
21.43—Nov 2025
12
Kimi K2.5OPEN
20.56$2.85Feb 2026
13
GPT-5.1
19.83—Nov 2025
14
Gemini 2.5 Pro (Jun 2025)
17.69—Jun 2025
15
o3
16.3—Dec 2024
16
GPT-5 mini
15.38—Aug 2025
17
Gemini 2.5 Pro (Mar 2025)
14.03—Mar 2025
18
o4-mini
13.95—Apr 2025
19
Gemini 2.5 Pro (May 2025)
13.66—May 2025
20
Claude Sonnet 4.5
9.37—Sep 2025
21
Gemini 2.5 Flash (Apr 2025)
7.65—Apr 2025
22
Claude Opus 4.1
7.06—Aug 2025
23
Gemini 2.5 Flash (May 2025)
6.47—May 2025
24
Claude Opus 4
6.22—May 2025
25
Gemini 3.1 Flash-Lite
4.03$1.50Mar 2026
26
Claude 3.7 Sonnet
3.4—Feb 2025
27
o1
3.32—Dec 2024
28
Claude Sonnet 4
3.11—May 2025
29
Gemini 2.0 Flash Thinking (Jan 2025)
1.85—Jan 2025
30
Llama 4 MaverickOPEN
0.92—Apr 2025
31
GPT-4.5
0.67—Feb 2025
32
GPT-4.1
0.63—Apr 2025
33
Amazon Nova Pro
0—Dec 2024
34
Claude 3.5 Sonnet (October 2024)
0—Oct 2024
35
Mistral Medium 3
0—May 2025
36
GPT-4o (Nov 2024)
0—Nov 2024
37
Gemini 1.5 Pro (Sept 2024)
0—Sep 2024

Scores from Epoch AI, updated —. Cost is joined from our daily API pricing where a model matches; models we don't price still appear with their score. Always verify on the benchmark's own site before relying on a number.