PLSQLBench: leaderboard
Metric: Mean test pass@1 (%; mean of the three held-out test sets: MBPP+-PLSQL (308 tasks), Spider2 single-turn test (103) and Spider2 multi-turn test (63 conversations); direct generation of Oracle PL/SQL program units at temperature 0, executed on Oracle Autonomous Database 23ai; per-task fraction of unit tests passed by one solution, averaged over tasks). Source: arxiv.org. Saturation forecast: Around 2028. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 64.96 |
| 2 | Claude Opus 4.8 | 63.49 |
| 3 | Gemma 4 31B | 62.26 |
| 4 | GPT-5.6 Sol | 61.71 |
| 5 | GPT-5.4 Mini | 58.6 |
| 6 | Grok 4.3 | 56.92 |
| 7 | Gemini 2.5 Flash Lite | 50.39 |
| 8 | Llama 4 Maverick | 46.8 |
Interactive version: theaggregate.ai/benchmark?slug=plsqlbench · How It Works · Data refreshed daily, snapshot 2026-09-26.