Browser-Use
An open-source framework for AI agent browser automation, built on top of Playwright. It enables agents to programmatically control a browser to perform automated tasks such as web page navigation and form submission.
Timeline (last 90 days)
Browser-Use lets agents observe via the accessibility tree and screenshots, and execute actions through clicking, typing, and scrolling.
Browser-Use and Playwright can be installed via the commands pip install browser-use playwright and playwright install-deps.
The course's Browser-Use framework is based on the Playwright automation program at its core, enabling programmatic Agent control of the browser.
All Facts (7)
Browser-Use的技术架构分为三层：浏览器控制层、感知抽象层和AI决策层
90%Unverified在开源领域，与Browser-Use类似定位的项目还有LaVague、WebArena、Skyvern等
90%UnverifiedBrowser-Use系统通过DOM解析提取可交互HTML元素，为每个元素分配唯一编号并用彩色边框可视化标注，将结构化描述和截图发送给LLM进行决策
90%UnverifiedBrowser-Use的感知抽象层能将约50KB的原始DOM压缩到几KB的结构化文本
70%UnverifiedBrowser-Use lets agents observe via the accessibility tree and screenshots, and execute actions through clicking, typing, and scrolling.
50%UnverifiedBrowser-Use and Playwright can be installed via the commands pip install browser-use playwright and playwright install-deps.
50%UnverifiedThe course's Browser-Use framework is based on the Playwright automation program at its core, enabling programmatic Agent control of the browser.
50%