WebBrain
An AI browser extension that reads pages and automates tasks in Chrome, Firefox and Edge, with offline local models or cloud APIs.
Alternative to:

WebBrain is an AI browser agent for people who want to ask questions about pages or delegate web tasks from a sidebar. It works in your existing browser, including signed-in tabs, and starts in read-only Ask mode.
- Page reading: Summarize articles, answer questions and extract structured information from tables, links, forms and PDFs.
- Browser tasks: Act mode can click, type, fill forms, upload and download files, with permission prompts before consequential actions and optional plan approval before work starts.
- Reusable automation: Save successful runs as workflows you can repeat, export or share, and schedule tasks or watch pages for conditions that trigger work.
- Developer tools: Dev mode can inspect page source, styles, console output and network activity, and make reversible page edits.
- Coding agent access: A local MCP bridge lets Claude Code, Codex, Cursor and other MCP clients delegate tasks to signed-in Chromium tabs while retaining browser permission checks; the bridge doesn't support Firefox.
WebBrain supports Chrome, Firefox and Chromium browsers. Its Edge extension requires Windows. It runs in an open browser rather than headless, and it's free and open source under GPL-3.0-or-later, with no WebBrain account required.
Local models work through llama.cpp, Ollama, LM Studio or another OpenAI-compatible server. They can process page content on your machine and support offline use with locally available content. Cloud choices such as OpenAI, Claude and OpenRouter receive the content needed for model requests. The extension can pair a text model with a separate vision model for screenshots. Local models generally need a context window of at least 16k tokens; 8k works with the reduced Compact tool set.