Open the side panel, ask in plain language, and Browser Copilot reads the page you're on and takes action — clicking buttons, filling forms, scrolling, switching tabs, walking through whole flows — while you watch or stay hands-off.
Private by construction. No accounts, no telemetry, no cloud server.
Works with DeepSeek, OpenAI, OpenRouter, Moonshot, DashScope, or local Ollama / LM Studio. Any OpenAI-compatible endpoint.
Four modes from plain chat to full auto. Semi-auto shows every action for your approval first — you stay in control.
Record yourself doing a task, or assemble it on a canvas from 56 block types — clicking, filling, loops, conditions, AI-agent steps.
Set recurring tasks — every few minutes, daily, or on weekdays. Run unattended prompts or summarize your GitHub review queue while you sleep.
Turn repeatable know-how — review checklists, style guides, extraction formats — into one-tap instructions the model must follow.
No accounts, no telemetry, no cloud server. Your keys and data stay on your machine. Passwords are filled locally and never shown to the model.
Three steps. No sign-up required.
Grab the latest ZIP from GitHub Releases. Unpack it, open
chrome://extensions, enable Developer mode, and click "Load unpacked".
Open Settings in the side panel. Enter the API endpoint and key for your preferred model provider — DeepSeek, OpenAI, OpenRouter, or a local Ollama instance.
Open the side panel on any page and ask in plain language. Browser Copilot reads the page and takes action — or build a workflow to automate entire procedures.
Any OpenAI-compatible model, coding agents via MCP, and Feishu/Lark for team collaboration.
A standalone MCP server gives your coding agent a real browser toolbox.
Get notified in a group chat when a task finishes. DM the bot from your phone and watch it execute in the browser on your machine.
Free, open source, and private by design. Download now and start in under two minutes.