DataClawEval - PrestoSQL-Trino: leaderboard
DataClawEval gives agents industrial data-engineering tasks rebuilt from production workflows: they must inspect the data, write and debug a job, and materialize correct output tables. This board covers the tasks that run on PrestoSQL and Trino. Grading is deterministic, comparing the produced artifact row by row against live databases and adding a process score for exploration, execution efficiency and self-verification.
Metric: Rule-based task score (%; per task a weighted sum, usually 0.7 and 0.3, of the artifact score, checked row by row against live databases by task-specific deterministic scripts, and the process score for exploration, execution efficiency and self-verification, mean over the 12 PrestoSQL/Trino tasks; one run per task, every model in the same Tencent CodeBuddy agent scaffold). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 16 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | CodeBuddy + Gemini 3.5 Flash | 77.7 | |
| 2 | CodeBuddy + DeepSeek V4 Pro | 77.7 | |
| 3 | CodeBuddy + Kimi K2.6 | 76.2 | |
| 4 | CodeBuddy + Kimi K2.7 | 76.1 | |
| 5 | CodeBuddy + GLM 5.1 | 75.9 | |
| 6 | CodeBuddy + MiniMax M3 | 75.4 | |
| 7 | CodeBuddy + GPT 5.5 | 75.3 | |
| 8 | CodeBuddy + Claude Opus 4.8 | 74.7 | |
| 9 | CodeBuddy + DeepSeek V4 Flash | 74.4 | |
| 10 | CodeBuddy + Gemini 3.1 Pro | 72.8 |
Interactive version: theaggregate.ai/benchmark?slug=dataclaweval-prestosql-trino · How It Works · Data refreshed daily, snapshot 2026-11-13.