Browser Automation
Discover 29 open source Browser Automation repositories from GitHub, each with an analysis of what it does, key features, use cases and alternatives. Browser Automation projects here are most often combined with Automation, AI Agents and TypeScript. Last updated October 4, 2026.
29 repositories · updated October 4, 2026
UI-TARS-desktop: Control Computers with a Multimodal AI Agent
UI-TARS-desktop is a desktop GUI agent that uses vision-language models to operate computers through natural-language instructions. It offers local and remote computer and browser operators for users building or exploring AI-driven automation.
index: Automate Complex Tasks in a Web Browser
Index is a Python browser agent that uses reasoning-capable language models to carry out multi-step web tasks and return structured results. It offers a CLI, a Python interface, and a serverless API, but the repository is archived.
opendia: Connect AI Models to Your Browser
OpenDia is an MCP server and browser extension that lets compatible AI clients inspect and control your existing browser. It suits people who want local browser automation across their logged-in sessions, tabs, and pages.
youtube-transcripts-machine: Extract YouTube Video Transcripts
YTM turns a YouTube video URL into a timestamped transcript with clickable time links and export options. It suits people who need a readable, shareable transcript and are comfortable configuring cloud browser and AI credentials.
BrowserOS: Browse with AI Agents and Automate Web Tasks
BrowserOS is a Chromium-based project with two desktop browsers: one for AI agents to automate logged-in web tasks, and another with an AI assistant for people browsing. It connects to MCP-compatible agents and keeps session data on your machine.
neko: Stream a Shared Virtual Browser Over WebRTC
Neko runs a browser or other desktop application in a Docker container and streams it to users over WebRTC. It suits watch parties, collaborative sessions, and remote access to an isolated browser.
magentic-ui: Automate Browser and File Tasks with AI Agents
MagenticLite is an experimental agent for browser and local-file workflows, designed to work with smaller models. It supports human approval and browser sandboxing, and suits people who want supervised automation rather than unattended computer control.
brightdata-mcp: Give AI Agents Access to Web Data
Bright Data MCP connects MCP-compatible agents to web search, scraping, structured extraction, and remote browser automation. It suits teams that need current public-web data without managing proxies or browser infrastructure, using a Bright Data API token.
vibe-tools: Add AI Teammates and Skills to Coding Agents
vibe-tools is a TypeScript CLI that connects coding agents to AI models, web research, repository analysis, browser automation, and developer-service integrations. It suits agent-driven workflows that need capabilities beyond editing local code.
carbonyl: Browse the Web from a Terminal
Carbonyl runs a Chromium-based browser directly in a terminal, including over SSH and without a window server. It suits users who need modern web pages and media in a text-console environment, but is more resource-intensive to build than a lightweight text browser.
splinter: Automate Web Browser Interactions in Python
Splinter provides a Python API for automating browser interactions in web application tests. Use it to exercise pages through drivers such as Selenium, or when you want a higher-level interface that can still access Selenium directly.
donutbrowser: Manage Isolated Chromium Browser Profiles
Donut Browser is a desktop Chromium profile manager built around a privacy-focused anti-detect engine. It is aimed at people who need separate browser identities, per-profile proxies, and optional sync or automation integrations.