PLSQLBench: leaderboard
PLSQLBench (Oracle, 2026) evaluates whether LLMs can write executable Oracle PL/SQL programs, scored by execution-based unit tests at temperature 0. Mean test pass at one, averaged over the three held-out test sets (MBPP+, Spider2 single-turn and multi-turn), in percent; higher is better.
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 October 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 64.96 | #121 |
| 2 | Claude Opus 4.8 | 63.49 | #56 |
| 3 | Gemma 4 31B | 62.26 | #393 |
| 4 | GPT-5.6 Sol | 61.71 | #29 |
| 5 | GPT-5.4 Mini | 58.6 | #389 |
| 6 | Grok 4.3 | 56.92 | #332 |
| 7 | Gemini 2.5 Flash Lite | 50.39 | #958 |
| 8 | Llama 4 Maverick | 46.8 | #1105 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=plsqlbench · How It Works · Data refreshed daily, snapshot 2026-11-12.