Humanity's Last Exam No Tools (Opus 5 System Card): leaderboard
Humanity's Last Exam (HLE), a 2,500-question multimodal benchmark of expert-level questions, answered without tools; accuracy in percent, higher is better. Anthropic's runs from Table 8.1.A of the Claude Opus 5 System Card (July 2026); the September card's HLE re-run reports different values and is a separate board.
Metric: Accuracy (%). Source: www-cdn.anthropic.com. Saturation forecast: Around January 2028. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 56.5 | #20 |
| 2 | Claude Opus 5 (Max) | 56.3 | #16 (Claude Opus 5) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=humanity-s-last-exam-no-tools-opus-5-system-card · How It Works · Data refreshed daily, snapshot 2026-11-05.