OpenAI GPT-5.5 Launch - Humanity's Last Exam (with tools): leaderboard
Metric: Score (%). Source: openai.com. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Mythos Preview | 64.7 |
| 2 | Claude Mythos 5 | 64.5 |
| 3 | GPT-5.4 Pro (xHigh) | 58.7 |
| 4 | Claude Opus 4.8 | 57.9 |
| 5 | GPT-5.5 Pro | 57.2 |
| 6 | Claude Opus 4.7 | 54.7 |
| 7 | GPT-5.5 | 52.2 |
| 8 | GPT-5.4 | 52.1 |
| 9 | Gemini 3.1 Pro (Preview) | 51.4 |
| 10 | GPT-5.2 Pro | 50 |
| 11 | GPT-5.2 | 45.5 |
Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-5-5-launch-humanity-s-last-exam-with-tools · How It Works · Data refreshed daily, snapshot 2026-09-05.