The settings
- CPU: How much simultaneous work each server can handle. More vCPUs help when many people use your app at the same time.
- Memory (RAM): How much data your app can hold in memory. More RAM helps with large files, media, or data-heavy work.
- Max machines: Autoscale only. The upper limit of servers Replit can add when your app is busy. Your app never scales beyond this number, which caps your maximum cost.
What it costs
The pane shows two numbers as you configure:- Per machine: The hourly cost of one server with your chosen configuration.
- At max traffic: The hourly cost if your app is busy enough to use every machine up to your maximum.
How billing differs by deployment type
The same configuration bills differently depending on your deployment type:- Autoscale: You pay while requests are being served. When your app is idle, it scales to zero and costs nothing to run.
- Reserved VM: Your server is always on, so you pay a fixed monthly cost for the chosen configuration.
- Static: No machines run your site, so machine configuration doesn’t apply.
- Scheduled: You pay for the machine during each scheduled run.
Apply your changes
Publish your app again after changing the configuration. The new machine settings take effect with the new deployment.Next steps
- Deployment types: Choose how your app is hosted.
- Publishing costs: View publishing costs by deployment type.