Inference V2

Inference Field

Base Inference

class BaseInference(raw_response)

Base class for V2 inference objects.

Parameters:: raw_response (dict[str, Any])

file: InferenceFile: File info for the inference.

id: str: ID of the inference.

job: InferenceJob: Job the inference belongs to.

model: InferenceModel: Model info for the inference.

Base Response

Inference Active Options

class InferenceActiveOptions(raw_response)

Active options for the inference.

Parameters:: raw_response (dict[str, Any])

confidence: bool: Boost the precision and accuracy of all extractions. Calculate confidence scores for all fields, and fill their confidence attribute.

data_schema: DataSchemaActiveOptions: Data schema options provided for the inference.

polygon: bool: Calculate bounding box polygons for all fields, and fill their locations attribute.

rag: bool: Enhance extraction accuracy with Retrieval-Augmented Generation.

raw_text: bool: Extract the full text content from the document as strings, and fill the raw_text attribute.

text_context: bool: Whether the text context feature was activated. When this feature is activated, the provided context is used to improve the accuracy of the inference.

Inference File

class InferenceFile(raw_response)

Inference File info.

Parameters:: raw_response (dict[str, Any])

alias: str: Alias of the file.

mime_type: str: Mime type of the file.

name: str: Name of the file.

page_count: str: Number of pages in the file.

Inference Model

class InferenceModel(raw_response)

Inference model info.

Parameters:: raw_response (dict[str, Any])

id: str: ID of the model.

RAG Metadata

class RAGMetadata(raw_response)

Metadata about the RAG operation.

Parameters:: raw_response (dict[str, Any])

retrieved_document_id: str | None: The UUID of the matched document used during the RAG operation.

Raw Text

class RawText(raw_response)

Raw text extracted from the document.

Parameters:: raw_response (dict[str, Any])

pages: list[RawTextPage]: Pages of raw text content.

Raw Text Page

class RawTextPage(raw_response)

Raw text extracted from the page.

Parameters:: raw_response (dict[str, Any])

content: str: Content of the raw text.