---
title: "Create Deployment"
method: POST
path: "/v1/projects/{project_id}/deployments"
tags: ["Deployment"]
---

# Create Deployment

`POST /v1/projects/{project_id}/deployments`

This endpoint creates a deployment, which is an instance for running a
model and performing predictions.

To create a deployment, you only need to provide a unique name. Upon
successful creation, the endpoint will return an HTTP URL where you can
send requests for predictions (also known as inferences).

The deployment name cannot be changed after creation.

Here's a breakdown of the deployment creation options:

**Model** (Optional): You can either:

- Provide the model data during deployment creation, making it ready to use
immediately.
- Upload a model to the deployment later and then use it for predictions.


**Threshold** (Optional):

- Specify a custom threshold value for your
predictions.
- If not provided, the default value used during model training will be
applied.

## Path parameters

- `project_id` integer, required

## Query parameters

- `timeout` integer, nullable

## Headers

- `apikey` string, nullable

## Request body

- DeploymentInput
  - `name` string, required
  - `modelId` string, uuid4, nullable
  - `threshold` number, nullable

## Response `201`

Successful Response

- BaseResponseDeploymentItem
  - `data` DeploymentItem, required
    - `id` string, uuid4, required
    - `predictionUrl` string, required
    - `name` string, required
    - `modelId` string, uuid4, nullable
    - `threshold` number, nullable

## Other responses

- `422` — Validation Error

---

[API](https://skmtc.net/landing-ai/apis/landinglens-api.md) · [All operations](https://skmtc.net/landing-ai/apis/landinglens-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/landing-ai/landinglens-api/versions/73e50adb1306/schema)