DeepResearch Bench is a deep research proxy benchmark dataset released by the University of Science and Technology of China in 2025. The related paper results are:DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents", aims to reveal the true distribution of human's needs for in-depth research in different fields. The dataset contains 100 PhD-level research tasks, each carefully crafted by experts in 22 different fields, including:

• **Science and Technology:**Physics, Chemistry, Biology, Environmental Science and Engineering
• **Finance and Business:**Investing, Personal Finance, Marketing and Human Resources
• **software:**Topics related to software usage and the Internet
• **other:**Art and design, entertainment, history, industry, transportation, tourism, etc.

Citation

@article{du2025deepresearch,
author    = {Mingxuan Du and Benfeng Xu and Chiwei Zhu and Xiaorui Wang and Zhendong Mao},
title     = {DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents},
journal   = {arXiv preprint},
year      = {2025},
}