PatRe - Office Action Quality (OA-DP): leaderboard
Metric: Overall office action quality on a 1-10 scale: mean of the soundness, clarity, constructiveness, completeness and legal-style ratings of a Gemini-3.1-Flash-Lite judge at temperature 0, office action generation under direct prompting: the model writes the office action from the claims (and any preceding rebuttal) with no prior art, over PatRe (480 recent USPTO patent examination histories across all eight IPC sections, with office actions, rebuttals, claim versions and cited references), temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 Mini | 4.56 |
| 2 | Qwen 3.5 27B | 4.35 |
| 3 | Gemini 2.5 Flash | 4.26 |
| 4 | DeepSeek V3.2 | 4.26 |
| 5 | Qwen 3.5 9B | 4.07 |
| 6 | Gemma 3 27B (IT) | 3.75 |
| 7 | GPT-4o Mini | 3.75 |
| 8 | Gemma 3 12B (IT) | 3.68 |
| 9 | Llama 3.3 70B Instruct | 3.45 |
| 10 | Llama 3.1 8B Instruct | 3.12 |
Interactive version: theaggregate.ai/benchmark?slug=patre-office-action-quality-oa-dp · How It Works · Data refreshed daily, snapshot 2026-10-07.