Create an ML Model

Create a new ML model within an ML model service

POST /v1/mlmodels

POST /v1/mlmodels

Python

from metadata.sdk import configure
from metadata.sdk.entities import MLModels
from metadata.generated.schema.api.data.createMlModel import CreateMlModelRequest

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

request = CreateMlModelRequest(
    name="customer_segmentation",
    displayName="Customer Segmentation Model",
    service="mlflow_svc",
    algorithm="KMeans",
    description="Segments customers into groups based on purchasing behavior",
    mlFeatures=[
        {
            "name": "total_spend",
            "dataType": "numerical"
        },
        {
            "name": "purchase_frequency",
            "dataType": "numerical"
        }
    ],
    mlHyperParameters=[
        {"name": "n_clusters", "value": "5"},
        {"name": "max_iter", "value": "300"}
    ],
    target="customer_segment"
)

model = MLModels.create(request)
print(f"Created: {model.fullyQualifiedName}")

Response

{
  "id": "6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2",
  "name": "customer_segmentation",
  "fullyQualifiedName": "mlflow_svc.customer_segmentation",
  "displayName": "Customer Segmentation Model",
  "description": "Segments customers into groups based on purchasing behavior",
  "algorithm": "KMeans",
  "mlFeatures": [
    {
      "name": "total_spend",
      "dataType": "numerical"
    },
    {
      "name": "purchase_frequency",
      "dataType": "numerical"
    }
  ],
  "mlHyperParameters": [
    {"name": "n_clusters", "value": "5"},
    {"name": "max_iter", "value": "300"}
  ],
  "target": "customer_segment",
  "version": 0.1,
  "updatedAt": 1769982669247,
  "updatedBy": "admin",
  "service": {
    "id": "ca22d46e-81b9-4e48-85b5-0adc44980da9",
    "type": "mlmodelService",
    "name": "mlflow_svc",
    "fullyQualifiedName": "mlflow_svc",
    "deleted": false
  },
  "serviceType": "Mlflow",
  "href": "http://localhost:8585/api/v1/mlmodels/6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2",
  "deleted": false,
  "owners": [],
  "tags": [],
  "followers": [],
  "votes": {
    "upVotes": 0,
    "downVotes": 0
  },
  "domains": []
}

Create an ML Model

Create a new ML model within an ML model service.

Body Parameters

name · string · required

Name of the ML model. Must be unique within the parent ML model service.

service · string · required

Fully qualified name of the parent MlModelService (e.g., mlflow_svc).

algorithm · string

Algorithm used by the ML model (e.g., KMeans, RandomForest, XGBoost, Neural Network).

displayName · string

Human-readable display name for the ML model.

description · string

Description of the ML model in Markdown format.

mlFeatures · array

Array of ML features used by the model.

Show properties

name · string

Name of the feature.

dataType · string

Data type of the feature (e.g., numerical, categorical).

featureSources · array

Sources of the feature data.

Show properties

name · string

Name of the feature source.

dataType · string

Data type of the source.

dataSource · object

Reference to the source entity (e.g., a table or column).

mlHyperParameters · array

Array of hyperparameters used by the model.

Show properties

name · string

Name of the hyperparameter.

value · string

Value of the hyperparameter.

target · string

Target variable or objective of the ML model.

server · string

Endpoint URL for the model serving server.

dashboard · object

Reference to an associated dashboard entity.

Show properties

id · string

UUID of the dashboard entity.

type · string

Type of entity (always dashboard).

owners · array

Array of owner references (users or teams) to assign to the ML model.

Show properties

id · string

UUID of the owner entity.

type · string

Type of owner entity (e.g., user, team).

name · string

Name of the owner entity.

domain · string

Fully qualified name of the domain to assign for governance purposes.

tags · array

Array of classification tags to apply to the ML model.

Show properties

tagFQN · string · required

Fully qualified name of the tag.

labelType · string

Type of label (e.g., Manual, Derived, Propagated).

state · string

State of the tag (e.g., Suggested, Confirmed).

extension · object

Custom property values defined by your organization’s metadata schema.

POST /v1/mlmodels

Python

from metadata.sdk import configure
from metadata.sdk.entities import MLModels
from metadata.generated.schema.api.data.createMlModel import CreateMlModelRequest

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

request = CreateMlModelRequest(
    name="customer_segmentation",
    displayName="Customer Segmentation Model",
    service="mlflow_svc",
    algorithm="KMeans",
    description="Segments customers into groups based on purchasing behavior",
    mlFeatures=[
        {
            "name": "total_spend",
            "dataType": "numerical"
        },
        {
            "name": "purchase_frequency",
            "dataType": "numerical"
        }
    ],
    mlHyperParameters=[
        {"name": "n_clusters", "value": "5"},
        {"name": "max_iter", "value": "300"}
    ],
    target="customer_segment"
)

model = MLModels.create(request)
print(f"Created: {model.fullyQualifiedName}")

Response

{
  "id": "6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2",
  "name": "customer_segmentation",
  "fullyQualifiedName": "mlflow_svc.customer_segmentation",
  "displayName": "Customer Segmentation Model",
  "description": "Segments customers into groups based on purchasing behavior",
  "algorithm": "KMeans",
  "mlFeatures": [
    {
      "name": "total_spend",
      "dataType": "numerical"
    },
    {
      "name": "purchase_frequency",
      "dataType": "numerical"
    }
  ],
  "mlHyperParameters": [
    {"name": "n_clusters", "value": "5"},
    {"name": "max_iter", "value": "300"}
  ],
  "target": "customer_segment",
  "version": 0.1,
  "updatedAt": 1769982669247,
  "updatedBy": "admin",
  "service": {
    "id": "ca22d46e-81b9-4e48-85b5-0adc44980da9",
    "type": "mlmodelService",
    "name": "mlflow_svc",
    "fullyQualifiedName": "mlflow_svc",
    "deleted": false
  },
  "serviceType": "Mlflow",
  "href": "http://localhost:8585/api/v1/mlmodels/6b04e1d8-b66d-4f78-ab21-beb5be2cf4f2",
  "deleted": false,
  "owners": [],
  "tags": [],
  "followers": [],
  "votes": {
    "upVotes": 0,
    "downVotes": 0
  },
  "domains": []
}

---

Returns

Returns the created ML model object with all specified properties and system-generated fields.

Response

id · string

Unique identifier for the ML model (UUID format).

name · string

ML model name.

fullyQualifiedName · string

Fully qualified name in format service.modelName.

displayName · string

Human-readable display name.

description · string

Description of the ML model in Markdown format.

algorithm · string

Algorithm used by the ML model.

mlFeatures · array

Features used by the ML model.

Show properties

name · string

Name of the feature.

dataType · string

Data type of the feature.

featureSources · array

Sources of the feature data.

mlHyperParameters · array

Hyperparameters used by the ML model.

Show properties

name · string

Name of the hyperparameter.

value · string

Value of the hyperparameter.

target · string

Target variable or objective of the ML model.

service · object

Reference to the parent ML model service.

Show properties

id · string

UUID of the ML model service.

type · string

Type of entity (always mlmodelService).

name · string

Name of the ML model service.

fullyQualifiedName · string

Fully qualified name of the ML model service.

serviceType · string

Type of ML model service (e.g., Mlflow, Sklearn, SageMaker).

owners · array

List of owners assigned to the ML model.

Show properties

id · string

UUID of the owner entity.

type · string

Type of owner entity (e.g., user, team).

name · string

Name of the owner entity.

domains · array

Domain assignments for governance.

tags · array

Classification tags applied to the ML model.

Show properties

tagFQN · string

Fully qualified name of the tag.

labelType · string

Type of label (e.g., Manual, Derived, Propagated).

state · string

State of the tag (e.g., Suggested, Confirmed).

extension · object

Custom property values defined by your organization’s metadata schema.

version · number

Version number for the entity (starts at 0.1).

---

Create or Update (PUT)

Use PUT /v1/mlmodels instead of POST to perform an upsert. If an ML model with the same fullyQualifiedName already exists, it will be updated; otherwise, a new ML model is created. The request body is the same as POST.

curl -X PUT "{base_url}/api/v1/mlmodels" \
  -H "Authorization: Bearer {access_token}" \
  -H "Content-Type: application/json" \
  -d '{ ... same body as POST ... }'

PUT will not return a 409 conflict error if the entity already exists — it will update the existing entity instead.

---

Bulk Create or Update (PUT)

Use PUT /v1/mlmodels/bulk to create or update multiple ML models in a single request. The request body is an array of create request objects.

curl -X PUT "{base_url}/api/v1/mlmodels/bulk" \
  -H "Authorization: Bearer {access_token}" \
  -H "Content-Type: application/json" \
  -d '[
    { "name": "model_one", "service": "mlflow_svc", "algorithm": "KMeans" },
    { "name": "model_two", "service": "mlflow_svc", "algorithm": "RandomForest" }
  ]'

---

Error Handling

Code	Error Type	Description
400	BAD_REQUEST	Invalid request body or missing required fields
401	UNAUTHORIZED	Invalid or missing authentication token
403	FORBIDDEN	User lacks permission to create ML models
409	ENTITY_ALREADY_EXISTS	ML model with same name already exists in service (POST only)