{"version":1,"type":"story","url":"https://digestai.news/story/researchers-release-webanything-framework-for-automated-web-agent-trai","json":"https://digestai.news/story/researchers-release-webanything-framework-for-automated-web-agent-trai.json","markdown":"https://digestai.news/story/researchers-release-webanything-framework-for-automated-web-agent-trai.md","slug":"researchers-release-webanything-framework-for-automated-web-agent-trai","headline":"Researchers release WebAnything framework for automated web agent training","summary":"Researchers introduced **WebAnything**, a fully automated framework that recreates functional, reusable web environments from real URLs without manual setup. The system uses an **Explorer-Builder** to capture browser-observable structures and interactions, while an independent **verifier** ensures accuracy and guides repairs if needed. It supports **244 website artifacts** across **14 domains** and **53 categories**, enabling **12,984 executed and accepted tasks**—more than human-assisted alternatives like **WebHarbor** in a blind audit (12 wins, two ties, no losses).\n\nFine-tuning **Qwen3.5-9B** on **5,856 interaction trajectories** improved task success rates by **9.9%** (REAL), **6.25%** (Odysseys-Easy Rubric Avg), **6.32%** (Task Avg), and **13.33%** (WebArena-Verified Overall). The code is available on GitHub.","keyPoints":["WebAnything automates web environment reconstruction from real URLs without manual per-site engineering","Supports 244 website artifacts across 14 domains and 53 categories, enabling 12,984 executed tasks","Fine-tuning Qwen3.5-9B on 5,856 trajectories boosted task success rates by up to 13.33%"],"whyItMatters":"WebAnything eliminates manual web environment setup, scaling agent training for real-world tasks. Its automated fidelity checks and performance gains could accelerate research on web-based AI agents.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":[],"models":["Qwen3.5-9B"],"people":[]},"firstPublishedAt":"2026-10-30T04:00:00Z","updatedAt":"2026-10-30T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"WebAnything: From the Wild Web to Scalable Agent Environments","url":"https://arxiv.org/abs/2610.37786","publishedAt":"2026-10-30T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Researchers release WebAnything framework for automated web agent training\", 30 October 2026, https://digestai.news/story/researchers-release-webanything-framework-for-automated-web-agent-trai","publisher":"Digest AI","title":"Researchers release WebAnything framework for automated web agent training","datePublished":"2026-10-30T04:00:00Z","url":"https://digestai.news/story/researchers-release-webanything-framework-for-automated-web-agent-trai"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}