v1/tools/agentic-document-analysis endpoint, the agentic-doc Python library, and the DPT-1 and dpt-2-mini parsing models are deprecated. Requests that use the legacy endpoint or a deprecated model return an error.
Legacy ADE Endpoint
The legacy Agentic Document Extraction endpoint (v1/tools/agentic-document-analysis) was the original endpoint for document processing. It combined document parsing and field extraction into a single API call.
This endpoint has been replaced with separate, function-specific APIs:
- ADE Parse API: Converts documents into structured Markdown with hierarchical JSON
- ADE Split API: Classifies and separates documents into sub-documents
- ADE Extract API: Extracts specific data fields from parsed documents
Legacy agentic-doc Library
The agentic-doc Python library is deprecated. Use the ade-python library for all new projects.Deprecated DPT-1 Parsing Model
DPT-1 was the original Document Pre-Trained Transformer (DPT) model for Agentic Document Extraction. It powered basic document parsing and was the only model available when Agentic Document Extraction launched. DPT-1 has been replaced by DPT-2, which builds on DPT-1 and adds support for complex tables, additional chunk types (such as logos, barcodes, and signatures), and improved layout detection. DPT-1 is deprecated. To migrate, update themodel parameter in your API calls or library code from dpt-1 to dpt-2. For more information, go to Document Pre-Trained Transformers (Parsing Models).
Deprecated DPT-2 Mini Parsing Model
Thedpt-2-mini model was a lightweight variant of DPT-2. It is deprecated.
To migrate, update the model parameter in your API calls or library code from dpt-2-mini to dpt-2.