Petri-Bench - Solved: leaderboard
LLM solvers probe simulated systems, infer hidden parameters and try to achieve target outcomes. Reports overall performance, solved cases, parameter recovery and individual simulation engines. Non-LLM search baselines remain outside the model comparison. This board reports Petri-Bench - Solved using Solved (%).
Metric: Solved (%). Source: huggingface.co. Saturation forecast: Around 2031. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 51.9 | #20 |
| 2 | GLM-5.2 | 50 | #80 |
| 3 | Gemini 3.1 Pro (Preview) | 48.9 | #64 |
| 4 | GLM-5.1 | 48.1 | #109 |
| 5 | Claude Sonnet 4.6 | 37.7 | #94 |
| 6 | GPT-5.5 | 36.7 | #33 |
| 7 | Claude Opus 4.8 | 36.7 | #39 |
| 8 | Claude Haiku 4.5 | 35.6 | #287 |
| 9 | GPT-5.6 Sol | 34.2 | #22 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=petri-bench-solved · How It Works · Data refreshed daily, snapshot 2026-11-05.