PageBolt Launches Browser Automation and Screenshot API with AI Voice Narration
PageBolt is a hosted REST API for screenshots, PDFs, AI-narrated video recording, and multi-step browser automation — with a native MCP server for Cursor, Claude Desktop, and Windsurf.
PageBolt, a hosted browser automation and web capture API, is now available to developers at pagebolt.dev. The platform provides eight production-ready endpoints — covering screenshots, PDF generation, OG image creation, multi-step browser sequences, AI-narrated video recording, page inspection, device presets, and quota tracking — behind a single API key.
Eight Tools, One API Key
PageBolt eliminates the need to run and maintain headless browsers in production. Developers make a single HTTP request and receive a rendered screenshot, a PDF, or a full narrated demo video without configuring Puppeteer, managing Chrome instances, or dealing with dependency updates.
The eight endpoints:
- Screenshot — Capture any URL with 25+ device presets, automatic cookie banner removal, ad blocking, and full-page support
- PDF — Generate print-ready PDFs from URLs or raw HTML
- OG Image — Create social preview images from templates or custom HTML
- Sequence — Execute multi-step browser flows (navigate, click, fill, screenshot) in a single browser session, returning multiple outputs
- Video — Record narrated demo videos with AI voice synchronized to each browser step using the Audio Guide feature
- Inspect — Return a structured map of page elements with CSS selectors, purpose-built for AI agent workflows
- Devices — List all supported device presets for emulation
- Usage — Check current quota consumption and plan limits
AI-Narrated Video Recording
The Video endpoint's Audio Guide feature lets developers define a script or per-step narration text. PageBolt records the browser session as an MP4 and overlays an AI voice reading each note as the corresponding step executes. The result is a polished, narrated demo video generated fully automatically — with no screen recording software, no manual editing, and no post-production.
"We built PageBolt because maintaining your own Puppeteer infrastructure is a surprisingly large tax on developer time," said a member of the PageBolt team. "The narrated video recording is the feature that keeps surprising people — you define your browser steps, add a sentence to each one, and get back a professional demo video. It hooks directly into GitHub Actions, so every pull request can have a live demo attached automatically."
MCP Server for AI IDEs
PageBolt ships an MCP (Model Context Protocol) server as the pagebolt-mcp npm package. Once installed, AI assistants in Cursor, Claude Desktop, and Windsurf can call PageBolt endpoints natively as tool calls — taking screenshots, generating PDFs, or recording demos without leaving the development environment.
Pricing
PageBolt offers a free tier with 100 requests per month, requiring no credit card. Paid plans start at $29/month (Starter, 5,000 requests), with Growth at $79/month (25,000 requests) and Scale at $199/month (100,000 requests). Starter and Growth plans include a 14-day free trial.
Availability
PageBolt is available now at pagebolt.dev. Full API documentation, code examples, and the pagebolt-mcp npm package are available on the website.
About PageBolt PageBolt is a developer-focused web capture and browser automation API. It provides hosted screenshot, PDF, video recording, and browser automation infrastructure so development teams can ship these features without managing headless browser infrastructure. PageBolt is headquartered online at pagebolt.dev.
Get Started Free
100 requests/month, no credit card
Screenshots, PDFs, video recording, and browser automation — no headless browser to manage.
Get Your Free API Key →