Replit’s deployment pricing is designed to scale with your app’s needs. Choose from usage-based billing that charges only when your app serves requests, or predictable flat-rate options for consistent workloads.

How billing works

All deployment costs are deducted from your monthly credits. You only pay usage-based fees after your monthly credits are fully used.
Credits apply automatically to all deployment costs. Unused credits don’t roll over to the next month. Learn more about usage-based billing.
For a hands-on understanding of deployment costs, explore our interactive pricing calculator.
Interactive deployment pricing calculator showing cost breakdown by deployment type

Deployment types

Choose the deployment type that best fits your app’s traffic patterns and resource needs.

Understanding request-based billing

With Autoscale Deployments, you only pay when your app is actively working. When no one visits your app, you pay nothing.
Autoscale Deployments use request-based billing—you’re charged only when your app serves traffic. Here’s how it works:
  1. App starts up when the first request arrives (if idle)
  2. Processes the request using compute resources
  3. Goes idle after 15 minutes of inactivity
Billing time: Often just 1-2 seconds per request, even for complex apps.

Request-based Billing Timeline

When no one visits your app, you pay nothing. When your app is busy, you pay for the compute resources used. Here’s a timeline of what happens when someone visits your app:

Request-based Billing Timeline

Instance
Running
Requests
Billable
StartedStopped
First, the server starts up. Then, it processes the requests. Finally, it goes idle. You only pay for CPU and memory during request processing.
When multiple requests arrive simultaneously (like the stacked blue bars), they share the same compute resources.Your billing time extends to cover all concurrent requests, but you don’t pay separately for each—just for the total time the server is working.
The gaps between green bars represent cost savings during idle time. At the end of the session, the server shuts down.

Compute units explained

Compute units measure the computational work your app performs:
  • CPU time: Processing power used (1 CPU second = 18 compute units)
  • Memory time: RAM consumed (1 GB-second = 2 compute units)
  • Duration: How long your app works on each request

Pricing breakdown

Cost examples by app type

These examples show realistic costs for different types of applications.

Monitor and control costs