PetriBench - MinimumTokenSteps (Hard): leaderboard
Metric: Exact-Match Accuracy (%). Source: arxiv.org. 36 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Grok 4.6 (High) | 76.8 |
| 2 | Grok 4.6 (xHigh) | 74 |
| 3 | Grok 4.5 (High) | 71.8 |
| 4 | Grok 4.6 (Medium) | 71 |
| 5 | Grok 4.5 (Medium) | 66.8 |
| 6 | Claude Opus 5 (xHigh) | 54.8 |
| 7 | GPT-5.6 Sol (xHigh) | 53.2 |
| 8 | Gemini 3.1 Pro (Preview) (High) | 50 |
| 9 | Claude Opus 5 (High) | 46.5 |
| 10 | Gemini 3.8 Flash (Medium) | 36.8 |
| 11 | Gemini 3.1 Pro (Preview) (Medium) | 35.5 |
| 12 | GPT-5.6 Sol (High) | 33.8 |
| 13 | Claude Opus 5 (Medium) | 26.8 |
| 14 | GPT-5.6 Terra (xHigh) | 18.8 |
| 15 | DeepSeek V4 Flash (High) | 17.2 |
Interactive version: theaggregate.ai/benchmark?slug=petribench-minimumtokensteps-hard · How It Works · Data refreshed daily, snapshot 2026-09-20.