SimulCost (Multi-Round): leaderboard
Metric: Success rate (%) averaged over the low, medium and high accuracy requirements, multi-round mode (up to 10 trials with simulation feedback, 2,304 tasks), on 11 physics simulators with analytically defined costs (fluid dynamics, heat transfer, solid mechanics, plasma physics); a task succeeds when the proposed parameters meet the accuracy requirement; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 80.8 | #91 |
| 2 | GPT-OSS-120B | 77.1 | #330 |
| 3 | Llama 3 70B Instruct | 70.6 | #624 |
| 4 | Qwen 3 32B | 67.7 | #424 |
| 5 | Claude 3.7 Sonnet (20250219) | 66.1 | #196 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=simulcost-multi-round · How It Works · Data refreshed daily, snapshot 2026-10-11.