RenoBench - Issue: leaderboard
Metric: Recall (%) of gold JATS issue annotations reproduced by the model's markup over 10,000 real-world references (lenient: fields inside valid tags still count when the outer XML is malformed); two-shot prompt; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-OSS-20B | 96.5 | #499 |
| 2 | Llama 3.1 8B Instruct | 96 | #1018 |
| 3 | Mistral 7B Instruct (v0.2) | 94.3 | #1310 |
| 4 | Qwen 3 0.6B | 92.7 | #1440 |
| 5 | Qwen 3 32B | 90.5 | #424 |
| 6 | Qwen 3 8B | 89.8 | #667 |
| 7 | Qwen 2.5 7B Instruct | 86.9 | #846 |
| 8 | Qwen 2.5 3B Instruct | 83.8 | #1138 |
| 9 | Gemma 3 1B (IT) | 64 | #1484 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=renobench-issue · How It Works · Data refreshed daily, snapshot 2026-10-11.