Getting Pupeteer to work in Replit?

Tips & Tricks
1

I’m trying to extract metadata from a Spotify podcast episode page (https://app.podfolio.me) and it’s not working. Here’s how Replit Agent described the problem:

Getting Puppeteer to work in the Replit environment is more complicated than expected for several reasons:

  1. Environment Constraints: Replit has limited resources and doesn’t include Chrome/Chromium by default. Puppeteer needs Chromium to run properly, which isn’t readily available in the Replit container environment.
  2. Sandbox Limitations: Replit’s security model restricts launching browser instances. Even with the correct flags like --no-sandbox, Chromium often can’t initialize properly because of the container restrictions.
  3. Memory Requirements: Headless browsers are resource-intensive. They require significant RAM (300-500MB minimum), which can exceed the limits of standard Replit environments.
  4. Dependency Issues: Installing Puppeteer automatically downloads Chromium (~170MB), but this download often fails in restricted environments like Replit, or the browser can’t be executed properly afterward.

Does this seem accurate?

2

Replit has all of those and you can use nix binaries for installing the chromium packages, that’s a hallucination by Agent

2 Likes
3

Thanks! I’ll try again!

1 Like
4

I went down a path of installing puppeteer and had to revert - after 10-12 checkpoints it still wasn’t working. It seems I can’t go back and review the chat after rolling back to determine what might have caused the loop, but I ended up outsourcing the work to an external API (PDF generation).

I used CloudConvert API when I couldn’t get puppeteer installed.