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

# Pull Model

> Initiates download of an Ollama model to the local Ollama instance

## Body Parameters

<ParamField body="modelName" type="string" required>
  Name of the model to pull (e.g., "llama2", "mistral:7b")
</ParamField>

<ParamField body="baseUrl" type="string">
  Optional Ollama server URL (defaults to [http://localhost:11434](http://localhost:11434))
</ParamField>

## Response

<ResponseField name="success" type="boolean">
  Indicates if the pull was initiated successfully
</ResponseField>

<ResponseField name="message" type="string">
  Status message about the pull operation
</ResponseField>

<RequestExample>
  ```bash cURL theme={null}
  curl --request POST \
    --url http://localhost:3000/api/v1/ollama/pull-model \
    --header 'Authorization: Bearer <token>' \
    --header 'Content-Type: application/json' \
    --data '{
      "modelName": "llama2"
    }'
  ```

  ```bash cURL (Custom Ollama URL) theme={null}
  curl --request POST \
    --url http://localhost:3000/api/v1/ollama/pull-model \
    --header 'Authorization: Bearer <token>' \
    --header 'Content-Type: application/json' \
    --data '{
      "modelName": "llama2:7b",
      "baseUrl": "http://192.168.1.100:11434"
    }'
  ```

  ```javascript JavaScript theme={null}
  const response = await fetch('http://localhost:3000/api/v1/ollama/pull-model', {
    method: 'POST',
    headers: {
      'Authorization': 'Bearer <token>',
      'Content-Type': 'application/json'
    },
    body: JSON.stringify({
      modelName: "llama2"
    })
  });

  const data = await response.json();
  ```

  ```python Python theme={null}
  import requests

  url = "http://localhost:3000/api/v1/ollama/pull-model"
  headers = {
      "Authorization": "Bearer <token>",
      "Content-Type": "application/json"
  }
  payload = {
      "modelName": "llama2"
  }

  response = requests.post(url, headers=headers, json=payload)
  data = response.json()
  ```
</RequestExample>

<ResponseExample>
  ```json 200 - Success theme={null}
  {
    "success": true,
    "message": "Started pulling model: llama2"
  }
  ```

  ```json 401 - Unauthorized theme={null}
  {
    "success": false,
    "error": "Unauthorized"
  }
  ```

  ```json 503 - Service Unavailable theme={null}
  {
    "success": false,
    "error": "Ollama is not running. Please start Ollama and try again."
  }
  ```

  ```json 500 - Server Error theme={null}
  {
    "success": false,
    "error": "Failed to pull model"
  }
  ```
</ResponseExample>

## Notes

<Warning>
  Ensure Ollama is running before attempting to pull a model. Start it with `ollama serve` or by running the Ollama app.
</Warning>

* Model pulls run asynchronously in the background
* Use the [Get Pull Status](/api-reference/ollama/pull-status) endpoint to monitor download progress
* Large models can take significant time to download
* Specify a tag (e.g., "llama2:7b") to pull a specific variant, or use just the name to pull the latest version

## Model Name Format

* `modelName`: Just the name (pulls latest) - e.g., "llama2"
* `modelName:tag`: Specific version - e.g., "llama2:7b", "mistral:latest"