Maxun

Maxun is an open source platform for building no-code “robots” that navigate websites like a real user and turn web content into structured data, clean markdown, or API outputs. It’s designed for quick web automation and repeatable extraction workflows, with options for both recorder-based and LLM-assisted extraction.

Key Features

• No-code recorder mode to capture browsing actions and reuse them as extraction robots
• LLM-powered extraction mode for describing desired fields in natural language
• Multiple robot types: extract structured data, scrape pages to markdown/HTML, crawl sites, and run automated web searches
• Generate REST-style endpoints from extraction robots to turn websites into structured APIs
• Scheduling for recurring runs and ongoing data collection
• Support for common dynamic patterns like pagination and infinite scroll
• Resilience features aimed at recovering from website layout changes
• SDK for programmatic control of robots and automation workflows

Use Cases

• Competitive and market research by tracking prices, listings, and product changes
• Lead generation and enrichment by extracting contact details and company data
• Feeding AI workflows with clean markdown content for RAG and document processing

Limitations and Considerations

• Web automation reliability can vary based on target site defenses (bot detection, CAPTCHAs) and frequent UI changes
• LLM-based extraction quality depends on the selected model and prompt context, and may require validation

Maxun fits teams that need repeatable web data collection without building custom scrapers from scratch, while still offering an SDK for deeper integration. It can scale from quick one-off extractions to scheduled pipelines that power internal systems and AI applications.