PageBolt is a hosted web capture and browser automation API built for developers and AI agent builders.
What Is It:
PageBolt provides a REST API with 8 endpoints that let you capture web content, automate browser interactions, and generate visual assets programmatically. It is designed as developer infrastructure — callable via HTTP, scriptable via CLI, and natively usable in AI IDEs via its MCP server.
How It Works:
PageBolt runs headless Chromium (via Puppeteer) plus ffmpeg on a hosted server. You send an API request with a URL and options; PageBolt loads the page in a real browser, performs the requested action, and returns a file or JSON result. No browser to manage, no infrastructure to maintain.
Use Cases:
- Automated screenshots for monitoring, OG image generation, or documentation
- PDF generation for invoices, reports, or archiving
- Multi-step browser automation sequences (fill forms, click, capture)
- Narrated video demos with AI voice for onboarding or sales
- Page inspection for AI agents that need structured element maps
Products:
- /screenshot — full-page or viewport screenshots
- /pdf — PDF from any URL
- /og-image — open graph images with custom dimensions
- /sequence — multi-step browser automation returning multiple outputs in one session
- /video — narrated screen recording with AI voice (Audio Guide)
- /inspect — element map for AI agent workflows
- /devices — browser device presets
- /usage — quota check
API Options:
Full REST API with API key authentication. Comprehensive docs at pagebolt.dev/docs.
MCP Server (pagebolt-mcp):
Native npm package — install once, use as a tool inside Claude Desktop, Cursor, Windsurf, and Cline without writing any integration code.
Pricing:
Free plan: 100 requests/month. Starter: $29/mo (5,000 req). Growth: $79/mo (25,000 req). Scale: $199/mo (100,000 req). 14-day free trials on Starter and Growth.