Migrate Stream Client - Arize AX Docs

The Stream Client is used for real-time logging of model predictions, providing lower latency than batch logging.

from arize.api import Client
client = Client(...)

from arize import ArizeClient
client = ArizeClient(...)

log()

The log() method migrates from client.log() to client.ml.log_stream().

Parameter Reference

This table provides a complete mapping of all parameters between v7 and v8, including which parameters were removed, renamed, or remain unchanged.

Parameter	v7	v8	Changes
`space_id`	Client init	Required per call	Must pass explicitly
`model_id`	Required	Required	Renamed to `model_name`
`model_type`	Required	Required	—
`environment`	Required	Required	—
`model_version`	Optional	Optional	—
`prediction_id`	Optional	Optional	—
`prediction_timestamp`	Optional	Optional	—
`prediction_label`	Optional	Optional	—
`actual_label`	Optional	Optional	—
`features`	Optional	Optional	—
`embedding_features`	Optional	Optional	—
`shap_values`	Optional	Optional	—
`tags`	Optional	Optional	—
`batch_id`	Optional	Optional	—
`prompt`	Optional	Optional	—
`response`	Optional	Optional	—
`prompt_template`	Optional	Optional	—
`prompt_template_version`	Optional	Optional	—
`llm_model_name`	Optional	Optional	—
`llm_params`	Optional	Optional	—
`llm_run_metadata`	Optional	Optional	—
`timeout`	N/A	✅ Optional	New parameter for request timeout

Side-by-Side Comparison

See the complete migration in action with this example showing both client initialization and streaming model predictions.

from arize.api import Client
from arize.utils.types import Environments, ModelTypes

# Client initialization
client = Client(
    api_key="your-api-key",
    space_id="your-space-id"
)

# Streaming a prediction
future = client.log(
    model_id="my-model",
    model_type=ModelTypes.BINARY_CLASSIFICATION,
    environment=Environments.PRODUCTION,
    model_version="v1.0",
    prediction_id="pred-123",
    prediction_timestamp=1609459200,
    prediction_label=1,
    features={"feature1": 0.5, "feature2": "value"},
    tags={"user_id": "user-456"},
    batch_id="batch-789"
)

# Get the result (blocks until complete)
response = future.result()

from arize import ArizeClient
from arize.ml.types import Environments, ModelTypes

# Client initialization
client = ArizeClient(api_key="your-api-key")

# Streaming a prediction
future = client.ml.log_stream(
    space_id="your-space-id",  # Now required per call
    model_name="my-model",  # Renamed from model_id
    model_type=ModelTypes.BINARY_CLASSIFICATION,
    environment=Environments.PRODUCTION,
    model_version="v1.0",
    prediction_id="pred-123",
    prediction_timestamp=1609459200,
    prediction_label=1,
    features={"feature1": 0.5, "feature2": "value"},
    tags={"user_id": "user-456"},
    batch_id="batch-789",
    timeout=30.0  # Optional, new parameter
)

# Get the result (blocks until complete)
response = future.result()