Skip to content

On demand servers

All server types found on Equinix Metalâ„¢ are eligible for on-demand deployment (data center availability permitting). Available inventory is not guaranteed for on-demand instance provisioning requests, see our documentation on capacity for more information about availability.

On-demand servers are charged a fixed-rate per hour. Billing begins as soon as the server is successfully provisioned, and ends when the server has been deleted from the user's project.

Server instances deployed on demand are never deleted by Equinix Metal without explicit permission from the instance owner, except in the event of account termination or indefinite restriction due to failure to comply with our Terms of Service.

Deploying an On Demand Server

You can deploy an on demand server from the Equinix Metal console, CLI, or API. You can also use any tooling or integrations that use them such as Terraform or Pulumi.

Using the Console

Log into your Equinix Metal account at and navigate to the Servers page. To spin up a new server, click + New Server. Next, choose On Demand.

You will be able to select your data center and, based on the capacity at the data center, your server type. The server options list the hardware type, price her hour and specs of each server class. Next, choose an Operating System. Depending on the hardware, Equinix Metal supports a large amount of OSes. Custom OS images and custom iPXE images are also supported.

If you want to deploy more than 1 server, you can adjust the number under Select Number. You can name your servers by changing the pre-filled hostnames. NOTE: You cannot use underscores in your hostname.

There are some optional settings if you need additional configuration of your servers.

Setting Description
User Data User Data allows you to specify a script or cloud config to perform various tasks after your server is ready.
Configure IPs Allows you to customize your IPv4 and IPv6 settings for your server.
Customize SSH Key Access Change which keys are added to your server. By default, every user's SSH key in your organization will be added to your new server.

You choices will be summarized in the Summary table. Once you are happy with the summary, you can deploy your server with the Deploy Now button.

Using the API

You can deploy a server with a single POST to the projects/{id}/devices endpoint of the Equinix Metal API.

curl -X POST \
-H "Content-Type: application/json" \
-H "X-Auth-Token: <API_TOKEN>" \
"{id}/devices" \
-d '{
    "facility": "<facility_code>",
    "plan": "<server_type>",
    "operating_system": "<os_code>"

Your project ID is required in the request path. The body parameters facility, plan, and operating_system are all required. Optional fields include hostname, ip_addresses, userdata, and many others are available to further configure your servers. A complete listing and explanation of all the fields available is in the API Reference.

Using the CLI

Similar to the API, you can deploy an on demand server with a single command in the Equinix Metal CLI.

packet device create --project-id <project_UUID> --plan <server_type> --facility <facility_code> --operating-system <os_code>

Your project ID is one of the required flags. The other required flags are --facility, --plan, and --operating_system. Optional flags include --hostname and --userdata. Many others are available to further configure your servers. A complete list of flags and their short codes is available in the CLI Reference.