protostatis/unbrowser

Overview

Single-binary, Chrome-free headless browser for LLM agents. Handles 80% of web pages at 1/50th the per-session cost of Chrome. Serves as a cheap first layer in a two-engine scraping stack.

Best for

Best for
Developers needing cheap, lightweight headless browsing for routine web scraping in AI pipelines

Use cases

• Scraping simple web pages for LLM training data
• Automated data extraction for AI agent workflows
• Lightweight headless browsing for automation pipelines

How to use

Install

pip  install pyunbrowser   # in a venv on python3.10+

Tools exposed

• text_main

Tested with

Claude Desktop, Claude Code, Cursor, Cline

Notes

Single-binary, Chrome-free headless browser for LLM agents. Handles 80% of web pages at 1/50th the per-session cost of Chrome. Serves as a cheap first layer in a two-engine scraping stack.

7 stars on GitHub. Last updated 2026-05-29. Licensed Apache-2.0.

Use cases

• Scraping simple web pages for LLM training data
• Automated data extraction for AI agent workflows
• Lightweight headless browsing for automation pipelines

Pros

• Extremely low per-session cost compared to Chrome
• Single binary with no Chrome dependency, saving resources
• Fast execution due to Rust implementation

Cons

• Covers only 80% of pages, requiring a fallback engine for complex sites
• May miss JavaScript-heavy or modern single-page applications
• Limited to simple scraping scenarios without full browser capabilities

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.