> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pioneer.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Pioneer projects API: deploy and manage model endpoints

> Organize Pioneer models into projects, deploy fine-tuned models to project endpoints, and run inference without managing training job IDs in your application.

Projects give you a way to organize your trained models and manage their deployments. You create a project, deploy a fine-tuned model to it, and then run inference against that project's endpoint. Each project maintains a deployment history so you can track which model version is live and roll back if needed.

<Note>
  After deploying a model, use `POST /projects/:project_id/inference` to run predictions through the project endpoint — no need to track training job IDs in your application code.
</Note>

***

## List projects

`GET /projects`

Returns all projects in your account.

```bash theme={null}
curl https://api.pioneer.ai/projects \
  -H "X-API-Key: YOUR_API_KEY"
```

**Response**

<ResponseField name="projects" type="object[]">
  Array of project objects.

  <Expandable title="project properties">
    <ResponseField name="id" type="string">
      Unique project ID.
    </ResponseField>

    <ResponseField name="name" type="string">
      Human-readable project name.
    </ResponseField>

    <ResponseField name="created_at" type="string">
      ISO 8601 creation timestamp.
    </ResponseField>
  </Expandable>
</ResponseField>

***

## Create a project

`POST /projects`

Creates a new project.

**Request body**

<ParamField body="name" type="string" required>
  A unique, human-readable name for the project.
</ParamField>

```bash theme={null}
curl -X POST https://api.pioneer.ai/projects \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "production-ner"
  }'
```

**Response**

<ResponseField name="id" type="string">
  The new project's unique ID. Use this in all subsequent project and deployment requests.
</ResponseField>

<ResponseField name="name" type="string">
  The project name.
</ResponseField>

***

## Delete a project

`DELETE /projects/:project_id`

Permanently deletes a project and its deployment history. Any deployed models associated with the project will no longer be accessible through the project inference endpoint.

<Warning>
  Deleting a project removes the deployment endpoint immediately. Make sure any clients using the project inference endpoint are updated before deleting.
</Warning>

**Path parameters**

<ParamField path="project_id" type="string" required>
  The project ID.
</ParamField>

```bash theme={null}
curl -X DELETE https://api.pioneer.ai/projects/YOUR_PROJECT_ID \
  -H "X-API-Key: YOUR_API_KEY"
```

Returns `204 No Content` on success.

***

## Deploy a model to a project

`POST /projects/:project_id/deployments`

Deploys a trained model to a project endpoint, making it available for inference. Creating a new deployment replaces the currently active model for that project.

**Path parameters**

<ParamField path="project_id" type="string" required>
  The project ID to deploy to.
</ParamField>

**Request body**

<ParamField body="training_job_id" type="string" required>
  The ID of the completed training job whose model you want to deploy. The training job must be in `complete` status.
</ParamField>

```bash theme={null}
curl -X POST https://api.pioneer.ai/projects/YOUR_PROJECT_ID/deployments \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "training_job_id": "YOUR_TRAINING_JOB_ID"
  }'
```

**Response**

<ResponseField name="deployment_id" type="string">
  Unique ID for this deployment.
</ResponseField>

<ResponseField name="status" type="string">
  Current deployment status. Values: `deploying`, `active`, `failed`.
</ResponseField>

<ResponseField name="training_job_id" type="string">
  The training job ID that was deployed.
</ResponseField>

***

## List deployment history

`GET /projects/:project_id/deployments`

Returns the full deployment history for a project, ordered from most recent to oldest. Use this to track which model versions have been deployed and when.

**Path parameters**

<ParamField path="project_id" type="string" required>
  The project ID.
</ParamField>

```bash theme={null}
curl https://api.pioneer.ai/projects/YOUR_PROJECT_ID/deployments \
  -H "X-API-Key: YOUR_API_KEY"
```

***

## Run inference on a project model

`POST /projects/:project_id/inference`

Runs a prediction against the currently active model deployed to this project. This endpoint abstracts away the underlying training job ID so your application only needs to reference the project.

**Path parameters**

<ParamField path="project_id" type="string" required>
  The project ID.
</ParamField>

**Request body**

<ParamField body="text" type="string" required>
  The input text to run predictions on.
</ParamField>

<ParamField body="schema" type="object">
  The output schema for encoder models. Accepts optional keys: `entities` (list of entity type strings), `classifications` (list of `{task, labels}` objects), `structures`, and `relations`. For decoder models, use `"task": "generate"` instead.
</ParamField>

```bash theme={null}
curl -X POST https://api.pioneer.ai/projects/YOUR_PROJECT_ID/inference \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Apple CEO Tim Cook announced the new MacBook at WWDC.",
    "schema": {
      "entities": ["person", "organization", "product", "event"]
    }
  }'
```