Data Agent Benchmark (Built-in ReAct Agent): leaderboard
Metric: Pass@1 (%). Source: ucbepic.github.io. Saturation forecast: Around December 2026. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 55.51 |
| 2 | Gemini 3 Pro | 46.63 |
| 3 | GPT-5 Mini | 36.63 |
| 4 | GPT-5.2 | 29.91 |
| 5 | Gemini 2.5 Flash | 10.49 |
Interactive version: theaggregate.ai/benchmark?slug=data-agent-benchmark-built-in-react-agent · How It Works · Data refreshed daily, snapshot 2026-09-26.