---
title: "Launch GPU instance"
method: POST
path: "/compute/v1/instances"
tags: ["Compute"]
---

# Launch GPU instance

`POST /compute/v1/instances`

Launches a new GPU compute instance with automatic SSH key generation. Supports mounting Case.dev Vaults as filesystems. Instance boots in ~2-5 minutes. Perfect for batch OCR processing, AI model training, and intensive document analysis workloads.

## Request body

- object
  - `name` string, required — Instance name
  - `instanceType` string, required — GPU type (e.g., 'gpu_1x_h100_sxm5')
  - `region` string, required — Region (e.g., 'us-west-1')
  - `vaultIds` string[] — Vault IDs to mount

## Response `201`

Instance launching

- object
  - `id` string
  - `name` string
  - `status` string
  - `instanceType` string
  - `region` string
  - `pricePerHour` string
  - `gpu` string
  - `specs` object
  - `vaults` unknown[]
    - unknown
  - `createdAt` string
  - `message` string

## Other responses

- `400` — Invalid parameters
- `401` — Invalid or missing API key
- `404` — Instance type, region, or vault not found
- `409` — Instance with this name already exists
- `500` — Failed to launch instance

---

[API](https://skmtc.net/casemark/apis/case-dev-api.md) · [All operations](https://skmtc.net/casemark/apis/case-dev-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/casemark/case-dev-api/versions/5b7e64e6d6f9/schema)