PAIR-Bench - Targeted Repair: leaderboard
Metric: Targeted repair success (%; share of the hidden tests in the hinted failure scenario that pass after the repair turn, averaged over turns and initially failing instances; higher is better). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3.2 | 71.59 |
| 2 | Gemini 2.5 Flash Lite | 59 |
| 3 | GPT-4o Mini | 56.14 |
| 4 | Qwen 3 Coder 30B A3B Instruct | 55.89 |
| 5 | Llama 3.3 70B Instruct | 53.5 |
| 6 | Ministral 3 14B | 49.71 |
| 7 | Mistral Small 3.2 | 49.31 |
| 8 | Gemma 3 27B | 42.56 |
Interactive version: theaggregate.ai/benchmark?slug=pair-bench-targeted-repair · How It Works · Data refreshed daily, snapshot 2026-09-29.