providerplaneai
    Preparing search index...

    Request payload for OCR and document extraction operations.

    interface ClientOCRRequest {
        context?: { metadata?: Record<string, unknown>; requestId?: string };
        extras?: Record<string, unknown>;
        file?: ClientFileInputSource;
        filename?: string;
        images?: ClientReferenceImage[];
        includeBoundingBoxes?: boolean;
        language?: string;
        mimeType?: string;
        model?: string;
        options?: Record<string, unknown>;
        prompt?: string;
        structured?: ClientOCRStructuredOptions;
    }

    Hierarchy (View Summary)

    Index

    Properties

    context?: { metadata?: Record<string, unknown>; requestId?: string }

    Execution context (tracing, requestId, etc.) Not sent to the provider.

    extras?: Record<string, unknown>

    Provider escape hatch for OCR-specific request parameters.

    Optional document or file source to process with OCR.

    This can be a local path, URL string, bytes, blob, file, or readable stream.

    filename?: string

    Optional filename hint when file is bytes or a stream.

    One or more images to process with OCR.

    includeBoundingBoxes?: boolean

    Whether the caller prefers region-aware OCR output when supported.

    language?: string

    Optional language hint for the document or image text.

    mimeType?: string

    Optional MIME type hint for the supplied file.

    model?: string
    options?: Record<string, unknown>

    Provider-agnostic tuning options (temperature, max tokens, etc.)

    prompt?: string

    Optional instruction to guide OCR behavior. Example: "Preserve line breaks and headings."

    Optional advanced structured extraction controls.