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

# ML Model Versions

> List and retrieve historical versions of an ML model

# ML Model Versions

Every change to an ML model entity creates a new version. Use these endpoints to view the version history and retrieve specific versions.

## List Versions

<ParamField path="id" type="string" required>
  UUID of the ML model.
</ParamField>

## Get Specific Version

Use `GET /v1/mlmodels/{id}/versions/{version}` to retrieve a specific version.

<ParamField path="id" type="string" required>
  UUID of the ML model.
</ParamField>

<ParamField path="version" type="string" required>
  Version number to retrieve (e.g., `0.2`).
</ParamField>

<RequestExample dropdown>
  ```python GET /v1/mlmodels/{id}/versions theme={null}
  from metadata.sdk import configure
  from metadata.sdk.entities import MLModels

  configure(
      host="https://your-company.open-metadata.org/api",
      jwt_token="your-jwt-token"
  )

  # List all versions
  versions = MLModels.get_versions("6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2")
  for v in versions:
      print(f"Version {v.version}: {v.description}")

  # Get a specific version
  model_v1 = MLModels.get_specific_version(
      "6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2",
      "0.1"
  )
  print(f"Original algorithm: {model_v1.algorithm}")
  ```

  ```java GET /v1/mlmodels/{id}/versions theme={null}
  import static org.openmetadata.sdk.fluent.MlModels.*;

  // List all versions
  var history = MlModels.getVersions("6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2");

  for (var version : history.getVersions()) {
      System.out.println("Version: " + version);
  }

  // Get a specific version
  var v1 = MlModels.getSpecificVersion(
      "6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2",
      0.1
  );
  System.out.println("Algorithm: " + v1.getAlgorithm());
  ```

  ```bash GET /v1/mlmodels/{id}/versions theme={null}
  # List all versions
  curl "{base_url}/api/v1/mlmodels/6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2/versions" \
    -H "Authorization: Bearer {access_token}"

  # Get a specific version
  curl "{base_url}/api/v1/mlmodels/6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2/versions/0.1" \
    -H "Authorization: Bearer {access_token}"
  ```
</RequestExample>

<ResponseExample>
  ```json Response (List Versions) theme={null}
  {
    "entityType": "mlmodel",
    "versions": [
      "{\"id\":\"6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2\",\"name\":\"customer_segmentation\",\"fullyQualifiedName\":\"mlflow_svc.customer_segmentation\",\"version\":0.2,\"description\":\"Updated: Customer segmentation using KMeans with 5 clusters\",\"algorithm\":\"KMeans\",\"serviceType\":\"Mlflow\"}",
      "{\"id\":\"6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2\",\"name\":\"customer_segmentation\",\"fullyQualifiedName\":\"mlflow_svc.customer_segmentation\",\"version\":0.1,\"algorithm\":\"KMeans\",\"serviceType\":\"Mlflow\"}"
    ]
  }
  ```
</ResponseExample>

***

## Returns

**List versions** returns an object with `entityType` and a `versions` array of serialized entity snapshots (newest first).

**Get specific version** returns the full ML model object as it existed at that version.

***

## Error Handling

| Code  | Error Type     | Description                             |
| ----- | -------------- | --------------------------------------- |
| `401` | `UNAUTHORIZED` | Invalid or missing authentication token |
| `404` | `NOT_FOUND`    | ML model or version does not exist      |