RecordHTML guides
Puppeteer vs. Playwright vs. timecut vs. RecordHTML
These tools do not all solve the same problem. Puppeteer and Playwright are browser automation building blocks, timecut is a specialized animation-capture project, and RecordHTML is a self-serve recording workflow with accounts and saved exports. Compare capture method and operational responsibility before comparing a feature list. The right choice depends on whether you need tests, custom deterministic rendering, a managed creative export, or an interactive walkthrough.
Puppeteer: browser control as a building block
Puppeteer offers programmatic browser control and documented screenshot and recording workflows. The current Page.record() documentation describes MP4 recording through Chrome DevTools Protocol; Page.screencast() is marked obsolete. Verify the API supported by your installed version. A browser API does not by itself provide your end-user upload flow, job queue, durable storage, retry policy or campaign delivery checks.
Before choosing it, verify the API and requirements for your installed version. Decide how you will handle process failures, browser resources, output encoding and cleanup. If you implement screenshot-based frame stepping, controlling one animation timeline does not necessarily control videos, CSS animations, randomness or asynchronous page state. Those dependencies are part of your implementation.
Playwright: recording in an automation and testing workflow
Playwright is designed for browser automation and testing, and its documentation includes video recording for browser contexts. Videos are saved when the context closes; the documented default video size scales the viewport to fit 800×800 unless configured. This can help diagnose test interactions, but should not be assumed to provide every codec, fractional frame rate or bitrate setting a campaign requests.
If you build a creative export service on top of it, document the capture and encoding stages separately. Verify the recorded area, output dimensions, lifecycle and file handling for the actual browser context. Add whatever encoding, storage and validation your delivery process requires rather than claiming that a video option automatically solves the whole workflow.
timecut: evaluate the specialized capture model
timecut documents a virtual-time approach using timeweb, timesnap, Puppeteer snapshots and FFmpeg encoding. Its README explicitly warns that changes driven by CSS transitions or animations may not render as intended because its time overrides target JavaScript and video playback. Test your own animation rather than assuming one capture method advances every browser timeline.
Before adopting a project in production, review its runtime requirements, dependency compatibility and maintenance activity directly in the repository. This guide does not claim a support guarantee or benchmark one tool as faster than another. A representative test creative and a reproducible environment are more useful than a generic performance ranking.
RecordHTML: a self-serve real-time export workflow
RecordHTML provides a form for hosted URLs and, when configured, creative ZIP uploads. Choose viewport dimensions, duration, output format and video frame rates including 29.97 and 59.94 fps. Finished MP4, WebM or GIF recordings are saved for preview, download and rerun. H.264 bitrate controls and the 4416×1656 preset address specific delivery needs.
The capture is real-time browser playback, not a claim of deterministic virtual-time rendering. An exact encoded frame count does not guarantee the creative rendered without stutter. Account limits and service capacity still apply. GIF has no audio, accepts whole-number rates from 1 to 30 fps, and is limited to 2,073,600 output pixels per frame.
Choose by evidence and responsibility
For test-run videos, begin with the automation framework already used by the team. For a custom frame-stepped renderer, evaluate the specialized capture model against every independent timeline and media dependency. For repeated creative exports without building a user-facing recording system, evaluate the self-serve workflow using your actual delivery brief.
Use the same permitted source creative for a fair comparison. Check output dimensions, timing, audio, motion, measured file size and repeatability; record the settings and versions. Include engineering and operations work in the decision, but avoid unsupported claims about speed, perfect frames or exclusive capabilities. No benchmark data or invented example outputs are presented here.
Sources and scope
Checked October 22, 2026. Destination specifications and third-party tools can change. Follow the linked documentation and your actual placement brief rather than treating this guide as network approval.