WebAnything: From the Wild Web to Scalable Agent Environments
Summary
WebAnything 是一个全自动框架，能从真实 URL 重建可复用、源绑定的 Web 环境，用于可扩展的智能体训练与评估。在 244 个网站环境上微调 Qwen3.5-9B，在 REAL、Odysseys 与 WebArena-Verified 基准上取得显著成功率提升。
View Cached Full Text
Cached at: 10/30/26, 03:15 AM
Paper page - WebAnything: From the Wild Web to Scalable Agent Environments
Source: https://huggingface.co/papers/2610.37786
Abstract
WebAnything automates the reconstruction of reusable, source-grounded web environments from real URLs, enabling scalable agent training and evaluation with verified fidelity.
Web agents require diverse, executable, and resettable website environments for reproducible training and evaluation. Human-assisted mirrors demand per-site engineering, while generative construction can diverge from specified real websites. Capture-and-replay preserves observations, with interaction coverage constrained by captured content and replay capabilities. We presentWebAnything, a fully automated framework that reconstructs source-grounded, reusable Web environments from real URLs without per-site manual implementation or repair after initial setup and authorization. AnExplorer-Buildercollects browser-observable structures, resources, and interaction outcomes; an independent verifier checksfrozen candidatesand guides local repair or targetedsource recapture. We construct 244 website artifacts across 14 application domains and 53 categories, supporting 12,984 executed and accepted tasks. Compared with human-assisted WebHarbor,WebAnythingachieves higher aggregate scores on three fidelity metrics, with 12 wins, two ties, and no losses in a blind audit. Fine-tuningQwen3.5-9Bon 5,856interaction trajectoriesimprovesREAL successby 9.9 percentage points, Odysseys-Easy Rubric Avg and Task Avg by 6.25 and 6.32 percentage points, andWebArena-VerifiedOverall success by 13.33 percentage points under the fixed harness. Code: https://github.com/china-qijizhifeng/Web-Anything
View arXiv pageView PDFProject pageGitHub1Add to collection
Get this paper in your agent:
hf papers read 2610\.37786
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2610.37786 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2610.37786 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2610.37786 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
WebAnything: From the Wild Web to Scalable Agent Environments
WebAnything presents a fully automated framework that reconstructs source-grounded, reusable web agent environments from real URLs, building 244 website artifacts supporting 12,984 tasks and showing that fine-tuning Qwen3.5-9B on its interaction trajectories improves web agent success rates across multiple benchmarks.
Weblica: Scalable and Reproducible Training Environments for Visual Web Agents
Apple Research introduces Weblica, a framework for creating scalable and reproducible training environments for visual web agents using HTTP caching and LLM-based synthesis.
Agent World Model in the Wild: Computer-Use Agent Environment Synthesis
WildAWM 提出一个可扩展的流水线，用于合成可控的计算机使用环境与任务来训练通用智能体，支持 GUI、MCP 和 CLI 交互，并在 8 个分布外基准上取得显著提升。
WebChallenger: A Reliable and Efficient Generalist Web Agent
WebChallenger is a new web agent framework that achieves strong performance across multiple benchmarks using open-weight models without fine-tuning, by replicating human cognitive advantages through architecture design rather than model scale.
Training Needs Trustworthy Worlds: Verified Synthetic Web Environments for Agent Learning
This paper introduces a framework for constructing verified synthetic web environments to improve the training of web agents, demonstrating enhanced performance and transferability across benchmarks.