Steps

Steps resource for step execution operations.

AsyncSteps

Asynchronous step execution operations.

Example:

>>> async with AsyncClient(api_key="...") as client:
    ...     status = await client.steps.run_async(
    ...         step_id="step_extraction",
    ...         file=Path("document.pdf")
    ...     )
    ...     result = await status.wait()
    ...     print(result.data)

Arguments:

client: The parent async client instance.

Methods:

get_status

def get_status(self, execution_id: str) -> StepExecutionStatus

Get the status of a step execution.

Arguments:

execution_id: The execution ID returned by run_async().

Returns:

The current execution status.

run_async

def run_async(self, step_id: str, file: FileInput | None = None, url: str | None = None, file_base64: str | None = None, content_type: str | None = None, document_metadata: dict[str, Any] | None = None) -> StepExecutionStatus

Execute a step asynchronously.

Arguments:

step_id: The ID of the step to execute. file: File to process (Path, bytes, or file-like object). url: URL of the document to process (alternative to file). file_base64: Base64-encoded document (alternative to file). content_type: Content type of the file. Auto-detected if not provided. document_metadata: Additional metadata to include with the document.

Returns:

The initial execution status with execution_id.

Properties:

• with_raw_response : Access methods that return raw HTTP responses.

Steps

Synchronous step execution operations.

Steps allow executing predefined document processing workflows.

Example:

>>> client = Client(api_key="...")
    >>> status = client.steps.run_async(
    ...     step_id="step_extraction",
    ...     file=Path("document.pdf")
    ... )
    >>> result = status.wait()
    >>> print(result.data)

Arguments:

client: The parent client instance.

Methods:

get_status

def get_status(self, execution_id: str) -> StepExecutionStatus

Get the status of a step execution.

Arguments:

execution_id: The execution ID returned by run_async().

Returns:

The current execution status.

Example:

>>> status = client.steps.get_status("exec_abc123")
    >>> if status.is_success():
    ...     print(status.data)

run_async

def run_async(self, step_id: str, file: FileInput | None = None, url: str | None = None, file_base64: str | None = None, content_type: str | None = None, document_metadata: dict[str, Any] | None = None) -> StepExecutionStatus

Execute a step asynchronously.

Initiates step execution and returns immediately with an execution ID. Use get_status() to poll for completion.

Arguments:

step_id: The ID of the step to execute. file: File to process (Path, bytes, or file-like object). url: URL of the document to process (alternative to file). file_base64: Base64-encoded document (alternative to file). content_type: Content type of the file. Auto-detected if not provided. document_metadata: Additional metadata to include with the document.

Returns:

The initial execution status with execution_id.

Raises:

ValueError: If no file input is provided. BadRequestError: If the request is invalid. NotFoundError: If the step doesn't exist.

Example:

>>> status = client.steps.run_async(
    ...     "step_invoice_extraction",
    ...     file=Path("invoice.pdf")
    ... )
    >>> print(f"Execution ID: {status.execution_id}")

Properties:

• with_raw_response : Access methods that return raw HTTP responses.