Deploy a GPU instance
create_instanceDestructive
Deploy a new GPU instance. Provide a name and either gpu (auto-picks cheapest) or type_id from list_gpu_types. Optionally attach an SSH key by ID for manual ssh -i access, and a storage volume by storage_id to boot with its files. For unattended or long-running work, set max_runtime_minutes and idle_shutdown_minutes: when either fires the instance is terminated (not paused), so copy results off before then. Once it is running, commands reach it through ssh_execute or the CLI (runcrate ssh ). If the environment has reached its monthly budget the launch fails: do not retry, tell the user.
Tool call template
Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {
"name": "<name>"
},
"name": "create_instance"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| gpu | No | GPU type for auto-deploy (e.g., A100, H100:2) | |
| name | Yes | Instance name | |
| region | No | Only launch in this region: a region from list_gpu_types, the start of one (us-east), or a city or country (canada). Nothing there fails and names the regions that have it; never launches elsewhere. Omit for any region | |
| type_id | No | Specific type ID from list_gpu_types | |
| template | No | OS template | |
| gpu_count | No | GPUs on the machine: 1, 2, 4 or 8 (default 1) | |
| workspace | No | Workspace ID (optional: defaults to the authenticated workspace) | |
| ssh_key_id | No | SSH key ID from list_ssh_keys (optional — keyless SSH always works without this) | |
| storage_id | No | Storage volume ID from list_volumes or create_volume (optional). Its files are copied to /workspace on the instance's local disk at boot; changes are copied back to the volume every few seconds and once more when the instance is deleted, so the files outlive the instance. Works with any GPU type and region. The volume must be in the same workspace and environment, active, and not attached to another running instance. Attach at launch only. | |
| environment | No | Environment name (optional: defaults to the environment this session acts in) | |
| include_community | No | Let a community host run it (off by default: datacenter machines only). A community host is run by an independent host, not a datacenter: the host operates the hardware and can access the machine, and it is not available on reserved terms. Only when the user asked for community hosts; needed to launch a community type_id. | |
| max_runtime_minutes | No | Terminate the instance automatically after this many minutes of running (5 to 10080). The clock starts when it first becomes running; an alert goes out 10 minutes before. Set it whenever you launch for unattended or long-running work, a bit above the expected job length (for a 2 hour job, 150). The instance is terminated, not paused. Omit for no limit. | |
| idle_shutdown_minutes | No | Terminate the instance when GPU utilisation stays under 5% with no SSH, web terminal or editor session for this many minutes (5 to 1440). GPU instances only. Set it whenever you launch for unattended or long-running work; 30 to 60 suits most jobs. The instance is terminated, not paused. Omit to leave idle stop off. |