# Create Deployment Request

Deploy an AI model onto a set of GPUs

## Properties

| Property | Type | Required | Description |
| --- | --- | --- | --- |
| `gpu-count` | integer | **yes** | Number of GPUs (1-8)<br/><br/>Min: `1`. |
| `gpu-type` | string | **yes** | GPU type family (e.g., gpua5000, gpu3080ti) |
| `model` | [Model reference. Provide either id or name.](/reference/api/_schemas/model-ref/) | **yes** |  |
| `name` | string | **yes** | Deployment name<br/><br/>Min length: `1`. |
| `replicas` | integer | **yes** | Number of replicas (>=1)<br/><br/>Min: `1`. |
| `inference-engine-parameters` | array[string] | no | Optional extra inference engine server CLI args |
| `inference-engine-version` | string | no | Allowed values: `0.12.0`, `0.15.1`, `0.16.0`, `0.17.0`, `0.18.0`, `0.18.1`, `0.19.0`, `0.19.1`, `0.20.0`, `0.20.1`, `0.20.2`, `0.21.0`, `0.22.0`, `0.22.1`.<br/><br/>Default: `0.22.1`. |