Proof of Time: leaderboard
Metric: Sample-Weighted Accuracy (%), 50-message agent budget. Source: arxiv.org. Saturation forecast: Around 2028. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 55.8 |
| 2 | Gemini 2.5 Pro | 53.6 |
| 3 | GPT-5.1 | 45 |
| 4 | Claude Opus 4.5 | 41.2 |
| 5 | GPT-5.2 | 40.1 |
| 6 | Gemini 3 Flash (Preview) | 39.2 |
| 7 | GPT-5 Mini | 37.7 |
| 8 | Gemini 2.5 Flash | 37.5 |
| 9 | Claude Haiku 4.5 | 35.1 |
| 10 | GPT-5 Nano | 33.9 |
| 11 | Claude Sonnet 4.5 | 31 |
Interactive version: theaggregate.ai/benchmark?slug=proof-of-time · How It Works · Data refreshed daily, snapshot 2026-09-25.