configuration

Instance Types

Available instance sizes, resource limits, and plan compatibility.

Every service runs on an instance type that determines how much CPU, RAM, and process capacity it gets. Choose the right size based on your workload.

Available Instances#

Type	CPU	RAM	Max PIDs	Price
Starter	0.5 CPU	512 MB	256	Free (included with plan)
Standard	1 CPU	2 GB	512	$7/mo per service
Pro	2 CPU	4 GB	1,024	$25/mo per service

Choosing an Instance#

Workload	Recommended Instance
Side projects, demos, personal blogs	Starter
Production APIs, medium-traffic sites	Standard
High-traffic services, data processing, ML inference	Pro
Databases with moderate load	Standard
Background workers, cron jobs	Starter or Standard

Plan Limits#

Your billing plan determines which instance types you can use:

Plan	Max Instance Type	Max Services	Max Databases
Hobby (free)	Starter	1	1
Starter ($5/mo)	Starter	3	1
Pro ($19/mo)	Standard	10	5
Business ($49/mo)	Pro	Unlimited	Unlimited

The instance type price is in addition to your plan price. For example, a Pro plan ($19/mo) with two Standard instances costs $19 + $14 = $33/mo.

Setting the Instance Type#

yaml

services:
  - name: api
    type: web-service
    instance: standard
    repo: https://github.com/alice/api

Resource Details#

CPU#

CPU is allocated in fractional cores. 0.5 CPU means half a core. Your service can burst above its limit briefly, but sustained usage is throttled to the allocated amount.

Memory#

RAM is a hard limit. If your service exceeds its memory allocation, the container is killed and restarted automatically (OOM kill). Monitor memory usage in the dashboard.

PIDs#

The PID limit caps the number of concurrent processes and threads. This prevents runaway process spawning. Most applications use well under 256 PIDs.

If your service keeps restarting, check the logs for OOM (out of memory) errors. You may need to scale up to a larger instance type.

Environment Variables

Custom Domains