Resource tiers
Every app runs on a size that determines its CPU and memory. Sizes let you right-size cost and performance — scale up for traffic spikes, down to save money.
The dashboard exposes five sizes, from the free tier up:
| Size | vCPU | Memory | CPU (millicores) | RAM |
|---|---|---|---|---|
g3.small |
0.25 | 512 MB | 250m | 512 MB |
g4.medium |
0.40 | 1 GB | 400m | 1024 MB |
g5.large |
0.70 | 2 GB | 700m | 2048 MB |
g6.xlarge |
1.40 | 4 GB | 1400m | 4096 MB |
g7.2xlarge |
2.10 | 6 GB | 2100m | 6144 MB |
g3.small is the free size — enough to try things out. Pick larger sizes for production
APIs, high-throughput services, and ML inference.
Choosing a size at deploy time
Section titled “Choosing a size at deploy time”In the dashboard’s Deploy App form, the Resources step lets you pick a size and how many instances to run.
Via the API, set cpu (millicores) and memory (MB) directly:
{ "cpu": 250, "memory": 512 } // g3.small{ "cpu": 700, "memory": 2048 } // g5.large{ "cpu": 2100, "memory": 6144 } // g7.2xlargeResizing without redeploying
Section titled “Resizing without redeploying”Edit the app’s settings from the dashboard, or PUT /apps with new cpu/memory values — the
app keeps running while the new size is applied. See
Configuration.
curl -X PUT https://api.cumin.dev/apps \ -H "Authorization: Bearer $TOKEN" \ -H "Content-Type: application/json" \ -d '{ "id": "<app-id>", "project_id": "…", "name": "my-api", "image": "docker.io/nginx:latest", "ports": [{ "name": "web", "number": 80 }], "cpu": 700, "memory": 2048, "instances": 1, "hibernated": false }'Sizes are priced per hour (see Billing). The free size costs nothing; larger sizes increase your hourly run-rate. Scale down (or hibernate an app) to cut spend instantly.