Pagecorder

The service records a web page as a hardware‑accelerated MP4 video through a simple API. Developers embed two JavaScript calls—window.pagecorder('start') and window.pagecorder('stop')—into their page; these calls are ignored by regular browsers, preserving normal functionality. After the page is prepared, a POST request to the /rapid/render endpoint submits the page URL or a ZIP archive containing the site, along with desired viewport dimensions. The API returns a job identifier, which can be polled via GET /rapid/render/:id until the recording completes, at which point the response includes the video URL, duration, and any diagnostic data.

For reliable, fast recordings, the documentation recommends packaging the entire site into a ZIP archive. The archive is served locally during rendering, eliminating network latency, preventing HTTP errors, and ensuring reproducible results regardless of external factors such as CDN performance or rate limits. This approach is suitable for static sites, animations, or any dynamic web content that can be captured within a defined time window.

The platform provides basic job lifecycle management, including timeout configuration, optional FTP upload destinations, and diagnostics like dropped frames or browser logs. It is positioned as an experimental developer tool for generating video representations of web pages without requiring manual screen capture.