---
title: "WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models"
license: CC-BY-4.0
license_url: https://creativecommons.org/licenses/by/4.0/
license_note: 'paper is arXiv non-exclusive distribution license (verified on arXiv abs page 2026-11-12); this pointer note is our own words and is CC BY 4.0; we link to the paper rather than reproduce it'
source: https://iinventedlight.com/library/notes/2401-13919-webvoyager.md
source_license: "arXiv non-exclusive distribution license"
attribution: 'Pointer by iinventedlight (AI-assisted). Source: Hongliang He, Wenlin Yao, Kaixin Ma, and 5 coauthors, WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models (arXiv non-exclusive distribution license). This note links to the paper rather than reproducing it.'
author: 'iinventedlight (AI-assisted)'
date: 2026-11-12
status: published
approved_by: "Cole Withington, 2026-11-12 14:44 PT, Librarian chat widget t51s316: \"Yes, publish all 100 second-list arXiv notes\""
checked_on: 2026-11-12
as_of: 2026-11-12
confidence: medium
untested: 'Pointer note only. Paper not reproduced. Abs page and license checked 2026-11-12.'
category: arxiv
topic_area: 'ai-research: agents'
tags: arxiv, agents, multimodal, web-agents, webvoyager
arxiv_id: 2401.13919
alias: 2401-13919-webvoyager
type: pointer
format: pointer
sources:
  - id: S1
    title: "WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models"
    publisher: arXiv
    authors: "Hongliang He and 7 coauthors (listed on the arXiv abs page)"
    url: https://arxiv.org/abs/2401.13919
    version_read: "v4 (6 Jun 2024)"
    accessed: 2026-11-12
    license: "arXiv non-exclusive distribution license"
---
# WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models

## Summary
[WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models](https://arxiv.org/abs/2401.13919) (2024) is an end-to-end multimodal web agent that carries out user instructions by interacting with real websites, rather than only with a simulator or a static snapshot.

What it introduced: a large multimodal model agent plus a new benchmark of tasks from 15 popular websites, and an automatic evaluation protocol that uses GPT-4V to score open-ended runs. The authors report a 59.1% task success rate, above both a GPT-4 all-tools baseline and a text-only WebVoyager variant, and they report 85.3% agreement between the automatic metric and human judges.

What it led to: a common reference for multimodal, in-the-wild web agents, used alongside WebArena, Mind2Web, and VisualWebArena. Live notes: [2307-13854-webarena](https://iinventedlight.com/library/notes/2307-13854-webarena/). Mind2Web and VisualWebArena are pointer notes in this batch; WebShop is a full note in batch F.

Where to read it: https://arxiv.org/abs/2401.13919 (v4, last revised 6 Jun 2024). Accepted to ACL 2024. Code and data: https://github.com/MinorJerry/WebVoyager.

The paper is posted on arXiv under the arXiv non-exclusive distribution license. This note links to it rather than reproducing it. The note itself is CC BY 4.0.

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Pointer by iinventedlight (AI-assisted). Paper license: arXiv non-exclusive distribution; this note links to it and does not reproduce it.

---

Agents: [/agents.md](/agents.md) | [/.well-known/agent-library.json](/.well-known/agent-library.json).
