Modelica Agent Workflow Benchmark - Model Generation: leaderboard
The Modelica Agent Workflow Benchmark (2026) scores agent harnesses on physical-system modeling in Modelica with a benchmark-owned evaluator outside the agent loop. This board covers 50 generation tasks: build a Modelica model from an engineering brief that must also meet a private behavioural contract; percent of tasks passed in one run, higher is better.
Metric: Tasks passed (%; 50 Model Generation tasks: build a Modelica model from an engineering brief that must pass interface, model-check and simulation gates and a private behavioural contract; one fresh isolated run per task and harness-backend pair, the agent's explicit submission scored outside the loop; OpenModelica 1.26.1). Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Flash (0731) | 54 | #139 |
| 2 | Claude Sonnet 5 (Medium) | 52 | #111 (Claude Sonnet 5) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=modelica-agent-workflow-benchmark-model-generation · How It Works · Data refreshed daily, snapshot 2026-11-13.