An AI web agent with a patched Firefox engine designed to avoid detection blocks.
Bypass anti-bot walls with a C++ patched Firefox engine and pluggable LLMs.
dots is an open-source web agent combining an OpenRouter LLM with a specialized, anti-detection Firefox browser. It suits professionals who need reliable web automation that won't get blocked by bot detection.
The tool pairs any model via OpenRouter with a C++ patched Firefox engine that masks its fingerprint at the engine level. It provides a live split-screen interface showing the chat on the left and the browser on the right, while supporting persistent profiles and proxy routing.
It suits developers and professionals building reliable browser automation workflows who are tired of standard scraping tools getting blocked. It does not suit those looking for a standard cloud-hosted SaaS with zero setup.
Engine-level fingerprinting prevents web pages from detecting WebDriver or automation flags.
Profile directories save cookies and login states between runs.
Pointer and key events mimic natural human movement rather than instant programmatic injection.
Expose the browser server to Claude Code, Codex, or Gemini CLI via invisible_playwright_mcp.
The software is free under the MIT licence, but you must provide your own OpenRouter API key to power the underlying AI model.
873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Yes, you can swap the model to any option available on OpenRouter using the command-line flag.
It uses a C++ patched Firefox engine where fingerprints are generated at the engine level, leaving no WebDriver or DevTools globals for pages to find.
Yes, the tool supports persistent profile directories that save your logins and cookies between runs.
See Zekai first in Google