PostEDA-Bench (ReAct) - DRC-Reasoning: leaderboard
Metric: Success rate (%; 30 residual post-flow DRC repair tasks, mean of 5 temperature-0 runs per task, ReAct agent). Source: arxiv.org. Saturation forecast: Around January 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 36.66 |
| 2 | GPT-5 | 31.33 |
| 3 | Gemma 4 31B (IT) (Thinking) | 27.99 |
| 4 | Qwen 3.5 27B | 27.32 |
| 5 | GPT-5 Mini | 20.66 |
| 6 | DeepSeek V3.2 | 16.66 |
| 7 | Qwen 3.5 122B A10B | 13.33 |
| 8 | Qwen 3.5 9B | 5.99 |
Interactive version: theaggregate.ai/benchmark?slug=posteda-bench-react-drc-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-25.