PAIR-Bench - Initial Fix: leaderboard
Metric: Initial fix rate (%; share of the 440 wrong-answer programs a model repairs to pass every hidden test in its first attempt, before any hint; higher is better). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3.2 | 65.68 |
| 2 | Gemini 2.5 Flash Lite | 53.18 |
| 3 | Qwen 3 Coder 30B A3B Instruct | 35.9 |
| 4 | GPT-4o Mini | 29.77 |
| 5 | Llama 3.3 70B Instruct | 27.72 |
| 6 | Ministral 3 14B | 22.95 |
| 7 | Mistral Small 3.2 | 13.4 |
| 8 | Gemma 3 27B | 10.22 |
Interactive version: theaggregate.ai/benchmark?slug=pair-bench-initial-fix · How It Works · Data refreshed daily, snapshot 2026-09-29.