HLCE: leaderboard
Humanity's Last Code Exam: 235 of the hardest competitive programming problems from ICPC World Finals and IOI (2010-2024). Measures Pass@1 averaged across both contest types.
Metric: Pass@1 (%). Source: github.com. Saturation forecast: Around January 2028. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O4 Mini (High) | 15.84 | #180 (O4 Mini) |
| 2 | Gemini 2.5 Pro (Preview 03-25) | 11.4 | #123 |
| 3 | O3 Mini (High) | 10.35 | #272 (O3 Mini) |
| 4 | O1 Mini | 5.95 | #323 |
| 5 | DeepSeek R1 | 4.83 | #251 |
| 6 | DeepSeek V3 (0324) | 3.53 | #340 |
| 7 | Claude 3.7 Sonnet | 2.48 | #249 |
| 8 | Claude 3.7 Sonnet (Thinking) | 2.12 | #249 (Claude 3.7 Sonnet) |
| 9 | Claude 3.5 Sonnet | 1.71 | #350 |
| 10 | GPT-4o (2024-05-13) | 1.22 | #325 |
| 11 | GPT-4o ChatGPT | 1.18 | |
| 12 | GPT-4o Mini | 0.48 | #599 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=hlce · How It Works · Data refreshed daily, snapshot 2026-10-20.