HiL-Bench Original Paper - SQL - Question Precision: leaderboard
Metric: Question Precision (%). Source: static.scale.com. 4 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.3 Codex | 55.3 |
| 2 | GPT-5.4 | 54.3 |
| 3 | Claude Opus 4.6 | 54.3 |
| 4 | Gemini 3.1 Pro (Preview) | 35.6 |
Interactive version: theaggregate.ai/benchmark?slug=hil-bench-original-paper-sql-question-precision · How It Works · Data refreshed daily, snapshot 2026-09-19.