Start here
Getting started
Create a project, deploy a public Docker image and get an HTTPS URL in a few minutes.
This guide takes you from a new account to a running container with a public HTTPS URL. It takes about five minutes; most of it is waiting for the first image pull.
1. Create an account and choose a plan #
Sign up with email, Google or GitHub. Creating projects and services needs a plan, so the dashboard asks you to start one first. The plan starts as soon as you subscribe — there is no free trial. See Billing & plans.
2. Create a project #
Open Projects in the dashboard and click New project. A project is an isolated environment with its own private network — use one per app or per environment (for example shop and shop-staging).
3. Deploy an image #
In the project, click New service and fill in:
| Field | Example | What it does |
|---|---|---|
| Name | web | Shown in the dashboard. The slug (the service's DNS name) is derived from it. |
| Image | nginx:1.28 | Any public image reference: name:tag, ghcr.io/owner/app:1.4.2 or name@sha256:…. |
| Public | on | Gives the service an HTTPS URL. Leave it off for databases and workers. |
| Port | 80 | The port your container listens on. Required for public services. |
| Mode | Always on | Or scale-to-zero for a public app that may sleep when idle. |
| Size | Small | CPU, memory and price — see sizes. |
Click Create. QuickTane checks the image against its registry, pins it to the exact digest and starts the rollout. The status shows deploying until the container is ready, then running.
4. Open your URL #
A public service gets an address of the form:
https://<service-slug>-<project-key>.quicktane.app
The project key is a short random part that is fixed for the life of the project, so the URL never changes when you redeploy. TLS is handled for you. The dashboard shows the URL on the service page, next to the logs, console and metrics.
Which images work #
- Public images only, from Docker Hub, GitHub (ghcr.io), Quay.io, GitLab, Amazon ECR Public, Microsoft (mcr.microsoft.com), Google (gcr.io), Google Artifact Registry.
- The image must have a
linux/amd64variant, and be at most 2 GiB compressed. - The image is pinned to a digest when you create or deploy it. Deploying the same reference again keeps the pinned digest, so tag each build uniquely (for example with the commit SHA) rather than reusing
latest. Redeploy on the service page (orPOST /v1/services/{id}/redeploy) checks the tag again and rolls out whatever it points to now. - Containers run in a sandboxed runtime (gVisor). Almost every image runs unchanged; a few that need special kernel features or privileged mode don't.
The same from the command line #
Create an API key with the services:manage scope, then:
curl -X POST https://api.quicktane.com/v1/services \
-H "Authorization: Bearer $QUICKTANE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"project_id": 12, "name": "web", "image": "nginx:1.28", "is_public": true, "port": 80}'
Next steps #
- Set environment variables and add a database from a template.
- Deploy new versions from CI with the REST API.
- Let your AI assistant manage services through the MCP server.
Something missing or wrong? Tell us — or open a ticket from the dashboard.