Skip to main content
The following information applies to the legacy Unstructured Partition Endpoint.Unstructured recommends that you use the on-demand jobs functionality in the Unstructured API instead. Unstructured’s on-demand jobs provide many benefits over the legacy Unstructured Partition Endpoint, including support for:
  • Production-level usage.
  • Multiple local input files in batches.
  • The latest and highest-performing models.
  • Post-transform enrichments.
  • All of Unstructured’s chunking strategies.
  • The generation of vector embeddings.
The Unstructured API also provides support for processing files and data in remote locations.
For the Unstructured Python SDK, the Unstructured Partition Endpoint returns errors primarily through the UnstructuredClientError class (the base class for all errors raised by the Unstructured Python SDK) and the HTTPValidationError class (inherited from UnstructuredClientError). Less common errors are returned through the following classes:
  • httpx.RequestError, the base class for request errors.
  • httpx.ConnectError, for HTTP connection request errors.
  • httpx.TimeoutException, for HTTP request timeout errors.
  • ServerError (inherited from UnstructuredClientError), for server-side errors.
  • ResponseValidationError (inherited from UnstructuredClientError), for type mismatches between the response data and the expected Pydantic model.
Each of the preceding classes has the following members:
MemberTypeDescription
messagestrThe eror message.
status_codeintThe HTTP response status code, for example 401.
headershttpx.HeadersA collection of HTTP response headers.
bodystrThe HTTP body. This can be an empty string if no body is returned.
raw_responsehttpx.ResponseThe raw HTTP response.
The following example shows how to handle the preceding errors. In this example, a required Unstructured API key is intentionally commented out of the code, so that a 401 Unauthorized error is intentionally thrown. 
import os
import json

from unstructured_client import UnstructuredClient
from unstructured_client.models.operations import PartitionRequest
from unstructured_client.models.shared import (
    PartitionParameters,
    Files,
    Strategy
)
from unstructured_client.models.errors import (
    UnstructuredClientError,
    HTTPValidationError
)
from unstructured_client.models.errors.servererror import ServerError
from unstructured_client.models.errors.responsevalidationerror import ResponseValidationError
import httpx

try:
    client = UnstructuredClient(
        # For example, intentionally leave out the API key to intentionally throw an error.
        # api_key_auth=os.getenv("UNSTRUCTURED_API_KEY")
    )

    filename = "PATH_TO_INPUT_FILE"

    request = PartitionRequest(
        partition_parameters=PartitionParameters(
            files=Files(
                content=open(filename, "rb"),
                file_name=filename,
            ),
            strategy=Strategy.VLM,
            vlm_model="gpt-4o",
            vlm_model_provider="openai",
            languages=['eng'],
            split_pdf_page=True,            # If True, splits the PDF file into smaller chunks of pages.
            # split_pdf_allow_failed=True,  # If True, the partitioning continues even if some pages fail.
            split_pdf_concurrency_level=15  # Set the number of concurrent request to the maximum value: 15.
        ),
    )

    response = client.general.partition(
        request=request
    )
    element_dicts = [element for element in response.elements]
    
    # Print the processed data's first element only.
    print(element_dicts[0])

    # Write the processed data to a local file.
    json_elements = json.dumps(element_dicts, indent=2)

    with open("PATH_TO_OUTPUT_FILE", "w") as file:
        file.write(json_elements)

except HTTPValidationError as e:
    print("Validation error (HTTP 422):", e)
except ServerError as e:
    print("Server error (HTTP 5XX):", e)
except ResponseValidationError as e:
    print("Response validation/type mismatch:", e)
except UnstructuredClientError as e:
    # This catches any other UnstructuredClientError not already caught above.
    # This and all of the other error classes in this example expose the following members:
    print("Other Unstructured client error:")
    print(f"Message:      {e.message}")
    print(f"Status code:  {e.status_code}")
    print(f"Body:         {e.body}")
    print(f"Raw response: {e.raw_response}")
    print(f"Headers:")

    for header in e.headers.raw:
        key = header[0].decode('utf-8')
        value = header[1].decode('utf-8')
        print(f"              {key}: {value}")

except httpx.ConnectError as e:
    print("HTTP connection error:", e)
except httpx.TimeoutException as e:
    print("HTTP timeout error:", e)
except httpx.RequestError as e:
    # This catches catch-all network errors from HTTP not already caught above.
    print("Other HTTPX request error:", e)
except Exception as e:
    # Optional: this catches any other unforeseen errors.
    print("Unexpected error:", e)
The results of running the preceding code are similar to the following: 
Message:      API error occurred: Status 401. Body: {"detail":"API key is missing, please provide an API key in the header."}
Status code:  401
Body:         {"detail":"API key is missing, please provide an API key in the header."}
Raw response: <Response [401 Unauthorized]>
Headers:
              date: <date-and-time-of-the-error>
              server: <server-identifier>
              content-length: 73
              content-type: application/json