Atlas

Benchmarks

← All benchmarks

Deep Research Bench — Find Dataset

Search · 2025-05-06

The DRB Find Dataset category asks an agent to identify links to datasets relevant to a query.

Top models (higher is better)

ModelScore
Opus 4.10.7
GPT-50.7
Grok 40.7
Opus 40.6
Opus 4.50.6
GPT-5.50.6
Opus 4.60.6
Gemini 3 Flash Preview0.6
Sonnet 40.6
Sonnet 4.50.6
Sonnet 4.60.6
Haiku 4.50.5
Opus 4.80.5
Gemini 2.5 Pro0.5
Gemini 3.1 Pro Preview0.5
GPT-5.10.5
GPT-5.20.5
Gemini 3.1 Flash-Lite Preview0.5
GPT-5.4 Mini0.5
o30.4
GPT-5.40.4
Gemini 3 Pro Preview0.4
Loading Atlas data…