Automate the Browser with Workers Browser Rendering API
summary
Cloudflare Workers’ Browser Rendering API runs a patched Puppeteer in a remote browser isolation service so serverless code can launch a Chromium instance, navigate pages, take screenshots and generate PDFs, and scrape or measure performance metrics. Demos show saving images and PDFs to R2 storage, running end-to-end checks, and collecting JS/CSS coverage. Current practical limit is short lived browsers (~60 seconds, two concurrent) and a demo uses Durable Objects to extend sessions by ~10s. API is open beta with a waitlist.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 5:46Cloudflare Workers
- 9:20Browser Rendering API
- 13:11Rendering API internals
- 16:43Browser automation use cases
- 18:51Screenshot demo with R2
- 25:23Durable Objects sessions
- 28:19PDF generation demo
- 33:20Performance metrics demo
- 36:37Future features
- 38:25Q&A
related talks
28mRethinking Browser Automation: Unleashing the Power of Puppeteer and Playwright
23mMaking Workers Work For You
38mThe Mysterious Life of Web Apps in the Background
41mYour Browser is Pooped: Using Web Workers To Create A Browser-Based Barcode Scanner
27mMaking Offline Suck Less with Service Workers
29mTyler Clark - React to Web Workers
43mService workers
28mWeb Workers " a graphical introduction"
