DRFlow (Basic Agent): leaderboard

Metric: Step F1 (%) on the 50-task original split of DRFlow (enterprise deep-research tasks in five domains: predict a personalized workflow of action steps from scattered heterogeneous sources): the task-wise harmonic mean of step recall and precision against the reference workflow (steps matched by an LLM judge), averaged over tasks; a minimal agent without a planning stage that ingests all documents and predicts the workflow directly; higher is better. Source: arxiv.org. Saturation forecast: Around 2031. 4 models tracked.

Top models

#ModelScore
1Claude Opus 4.513.54
2GPT-5.212.16
3DeepSeek V3.28.61
4Gemini 3.1 Pro (Preview)5.05

Interactive version: theaggregate.ai/benchmark?slug=drflow-basic-agent · How It Works · Data refreshed daily, snapshot 2026-09-29.