Exploring Autonomous Agentic Data Engineering: leaderboard
Metric: MATS Science (Scratch) (self-reported). Source: benchmarklist.com. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3.1 | 7.13 |
| 2 | Claude Sonnet 4 | 3 |
| 3 | DeepSeek R1 | 2.5 |
| 4 | Gemini 2.5 Pro | 2 |
| 5 | GPT-5.2 | 2 |
| 6 | Qwen 3 Max | 1.58 |
Interactive version: theaggregate.ai/benchmark?slug=exploring-autonomous-agentic-data-engineering · How It Works · Data refreshed daily, snapshot 2026-09-05.