PetriBench - Boundedness (Hard): leaderboard
Exact-match accuracy on PetriBench's boundedness questions at the Hard structural-complexity level, one of six task families over Petri-net state spaces. Reasoning effort is part of the row identity. Higher is better. Source: arxiv.org/abs/2609.19883.
Metric: Exact-Match Accuracy (%). Source: arxiv.org. Saturation forecast: Around 2032. 36 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol (xHigh) | 69 | #22 (GPT-5.6 Sol) |
| 2 | GPT-5.6 Sol (High) | 65 | #22 (GPT-5.6 Sol) |
| 3 | Claude Opus 5 (High) | 60 | #16 (Claude Opus 5) |
| 4 | Claude Sonnet 5 (Medium) | 58 | #65 (Claude Sonnet 5) |
| 5 | Gemini 3.8 Flash (High) | 55.5 | #30 (Gemini 3.8 Flash) |
| 6 | Claude Opus 5 (xHigh) | 54.5 | #16 (Claude Opus 5) |
| 7 | GLM-5 | 54 | #149 |
| 8 | Kimi K2.5 | 52.5 | #152 |
| 9 | Claude Sonnet 5 (High) | 52.5 | #65 (Claude Sonnet 5) |
| 10 | Claude Opus 5 (Medium) | 52.5 | #16 (Claude Opus 5) |
| 11 | Gemma 4 31B (High) | 52 | #212 (Gemma 4 31B) |
| 12 | Gemini 3.8 Flash (Medium) | 51.5 | #30 (Gemini 3.8 Flash) |
| 13 | Grok 4.6 (High) | 51 | #29 (Grok 4.6) |
| 14 | Claude Haiku 4.5 | 50.5 | #287 |
| 15 | Grok 4.6 (xHigh) | 50.5 | #29 (Grok 4.6) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=petribench-boundedness-hard · How It Works · Data refreshed daily, snapshot 2026-11-05.