How do I deploy my Replit app?

Also searched as: Replit app deployment · Replit app deployment cost · Deploy Replit app to Vercel · Deploy Replit app on AWS · Deploy Replit app for free

Updated 3 October 2026

Short answer

Inside Replit, open your project and use Publish. Replit calls making your app live "Publishing", and offers several deployment types: Autoscale (resources adjust with usage), Static (sites that don't change per user), Reserved VM (always on) and Scheduled. To host it somewhere else instead, export the code to a repository you own and deploy it like any other app. Either way, the keys you stored as Replit Secrets have to be set wherever it runs.

What's the problem

The app runs in Replit's editor. You want it at a real address, running when your laptop's closed, maybe on your own domain, maybe off Replit entirely. And you're not sure what it'll cost or what breaks when you move it.

Why it happens

  • The editor isn't the deployment. Running the app in the workspace is development. Publishing creates a separate, always-available version.
  • Different apps need different deployment types. An app with a backend and users needs a running server: Autoscale or Reserved VM. A plain website can be Static. Picking the wrong type leads to cold starts, cost surprises, or an app that doesn't work.
  • Settings live in Replit. Credentials kept as Replit Secrets don't travel with exported code. Elsewhere, the app starts without them.
  • Replit's conveniences don't exist elsewhere. Its database, storage and login services are part of the platform. Moving off Replit means replacing them, and other hosts expect your app to listen on the port they assign.

How to fix it

  1. To stay on Replit: select Publish, choose the deployment type (Autoscale for most apps with users, Static for plain sites), add a payment method if asked, and connect your custom domain.
  2. Check the published app works with production data, not just in the editor.
  3. To move off Replit, push the code to a GitHub repository you own.
  4. List every Secret in the project and set each one as an environment variable on the new host.
  5. Replace any Replit-specific services (database, storage, login) with standard equivalents, and migrate the data.
  6. Make the app listen on the host's port, usually from a PORT setting, on 0.0.0.0.
  7. Check security before real users arrive. See is my AI-built app secure.

When to call Preventionlabs

If Publish works and the app holds up, you're live. Call us when you want it off Replit and into infrastructure you own, and the move breaks things, or when the app works in the editor and nowhere else. A resurrection gets it running in your own hosting account with the user flows you list working end to end, protected to MVP level and on infrastructure that scales, then hands it back with notes on how to run and deploy it.

Submit your project for a free assessment

Free assessment. $10,000 AUD flat to get it live, only if we take it on and you go ahead.

Sources

  1. Replit: Publishingplatform docsThe action of making your app live is called “Publishing.”
  2. Replit: Publishing: deployment typesplatform docsAutoscale Deployment Automatically adjusts resources based on your app’s usage.
  3. Replit: Publishing: deployment typesplatform docsReserved VM Deployment Provides a consistent amount of computing resources for your app to run continuously.
  4. Replit: Secretsplatform docsStore credentials for external services as secrets and non-sensitive settings as configurations in your Replit project.
  5. Render: Web Services: Port bindinghosting platform docsEvery Render web service must bind to a port on host 0.0.0.0 to serve HTTP requests.
← All of Built with AI