Scaling and sizes

Choose a size, shared or dedicated vCPU, and how many instances Plexivo App Runner runs, with autoscaling and sleep on the Developer plan.

An app runs one or more identical instances. You choose the size of each instance and how many run. Plexivo App Runner adds and removes instances within the range you set.

Sizes#

Sizes come in three families. Each size is per instance.

SizeStandardHigh-CPUHigh-memory
nano0.125 vCPU · 256 MB
micro0.25 vCPU · 512 MB0.25 vCPU · 256 MB
small0.5 vCPU · 1 GB0.5 vCPU · 512 MB
medium1 vCPU · 2 GB1 vCPU · 1 GB1 vCPU · 4 GB
large2 vCPU · 4 GB2 vCPU · 2 GB2 vCPU · 8 GB
xlarge4 vCPU · 8 GB4 vCPU · 4 GB4 vCPU · 16 GB
2xlarge8 vCPU · 16 GB8 vCPU · 8 GB8 vCPU · 32 GB
4xlarge16 vCPU · 32 GB16 vCPU · 16 GB16 vCPU · 64 GB

The console lists the sizes you can choose. A size marked contact sales is one we set up for you. Prices for every size are on the pricing page.

Shared and dedicated vCPU#

  • Shared vCPU guarantees each instance a quarter of its vCPU and lets it use the full vCPU when it is free. It costs less.
  • Dedicated vCPU reserves every vCPU for the instance alone.

Developer apps run on shared vCPU. On Startup you choose shared or dedicated for each app when you create it. Business and Enterprise run on dedicated vCPU.

Instances and autoscaling#

On the app’s Scaling tab:

  • Instances is how many always run.
  • Autoscale up to is the most Plexivo may run. Set it equal to Instances for a fixed count.

Plexivo adds an instance when average CPU use passes 70% of the vCPU each instance has, and removes instances when it falls. Changing the count keeps the app serving; changing the size restarts it. Choose Apply changes, or Restart and apply for a new size.

Your plan sets how many instances an app may have: by default 1 on Developer, 4 on Startup, 10 on Business and 16 on Enterprise. The total vCPU and memory of everything in your organisation is also held to your plan. Both are in Settings → Limits & quotas.

If a bigger size does not fit where your project runs, the console says so before anything changes, and asks whether to move the project to where it fits.

With the API#

bash
# Price a shape first (optional)
curl -s "$API/deployments/$APP/shape/quote?cpu_m=1000&mem_mb=2048&replicas=2" \
  -H "Authorization: Bearer $PLEXIVO_TOKEN"

# 2 instances of 1 vCPU · 2 GB, autoscaling up to 4
curl -s -X PUT $API/deployments/$APP/shape \
  -H "Authorization: Bearer $PLEXIVO_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"cpu_m": 1000, "mem_mb": 2048, "replicas": 2, "max_replicas": 4}'

Sleep on the Developer plan#

A Developer app that has served no request for 15 minutes sleeps: it scales to zero. The next request starts it again. Its Overview shows App is sleeping while it rests. Apps on Startup and above never sleep.

Pause an app#

To stop an app without deleting it, choose Pause in the Danger zone of its Settings tab. Plexivo scales it to zero and stops billing its instances. Resume starts it again with the same address and settings.

How scaling is billed#

You pay for the instances that run. With autoscaling, the console shows the price as a range, from your minimum to your maximum. Each extra instance costs the same as the others and is charged only while it runs.

For businesses Set Instances to at least 2 for anything customer-facing, so one instance keeps serving while another restarts.

Next steps#

Updated 27 October 2026