# Towards a Realistic Long-Term Benchmark for Open-Web Research Agents

arXiv: https://arxiv.org/abs/2409.14913 · HTML: https://mlsanity.com/paper/2409.14913/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use.
- Builds on 6 of its 8 references that are in the map.
- First posted on arXiv 23 Sep 2024.

## Placed in

- LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/)

## Builds on (most cited first, 6 of 6)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents](https://mlsanity.com/paper/2207.01206/) (2022)
- [Mind2Web: Towards a Generalist Agent for the Web](https://mlsanity.com/paper/2306.06070/) (2023)
- [AdaPlanner: Adaptive Planning from Feedback with Language Models](https://mlsanity.com/paper/2305.16653/) (2023)
- [Evaluating Language-Model Agents on Realistic Autonomous Tasks](https://mlsanity.com/paper/2312.11671/) (2023)

## Built on by (most cited first, 4 of 4)

- [Deep Research Bench: Evaluating AI Web Research Agents](https://mlsanity.com/paper/2506.06287/) (2025)
- [Orak: A Foundational Benchmark for Training and Evaluating LLM Agents on Diverse Video Games](https://mlsanity.com/paper/2506.03610/) (2025)
- [GeoVista: Web-Augmented Agentic Visual Reasoning for Geolocalization](https://mlsanity.com/paper/2511.15705/) (2025)
- [Detecting Pipeline Failures through Fine-Grained Analysis of Web Agents](https://mlsanity.com/paper/2509.14382/) (2025)

## Abstract (arXiv)

We present initial results of a forthcoming benchmark for evaluating LLM agents on white-collar tasks of economic value. We evaluate agents on real-world "messy" open-web research tasks of the type that are routine in finance and consulting. In doing so, we lay the groundwork for an LLM agent evaluation suite where good performance directly corresponds to a large economic and societal impact. We built and tested several agent architectures with o1-preview, GPT-4o, Claude-3.5 Sonnet, Llama 3.1 (405b), and GPT-4o-mini. On average, LLM agents powered by Claude-3.5 Sonnet and o1-preview substantially outperformed agents using GPT-4o, with agents based on Llama 3.1 (405b) and GPT-4o-mini lagging noticeably behind. Across LLMs, a ReAct architecture with the ability to delegate subtasks to subagents performed best. In addition to quantitative evaluations, we qualitatively assessed the performance of the LLM agents by inspecting their traces and reflecting on their observations. Our evaluation represents the first in-depth assessment of agents' abilities to conduct challenging, economically valuable analyst-style research on the real open web.