Code-QA-Bench: leaderboard
Metric: Judge score (0-1; mean of accuracy, completeness and specificity; 528 code-derivable questions, documentation-stripped repository, GPT-5.4 judge). Source: arxiv.org. Saturation forecast: Around August 2027. 4 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4 Pro | 0.89 |
| 2 | Claude Opus 4.6 | 0.89 |
| 3 | Kimi K2.6 | 0.87 |
| 4 | Gemini 3.1 Pro (Preview) | 0.77 |
Interactive version: theaggregate.ai/benchmark?slug=code-qa-bench · How It Works · Data refreshed daily, snapshot 2026-09-25.