Skip to main content

AI Documents (PDF) Support

This module extends the AI Documents feature with PDF document support.

Features​

  • PDF Text Extraction: Extract text content from PDF documents
  • Page-by-Page Processing: Text is extracted from each page of the PDF

Getting Started​

  1. Enable the AI Documents (PDF) feature in Orchard Core admin
  2. Upload PDF files in the Documents tab of your chat interactions
  3. Text content will be automatically extracted and used for RAG

Technical Details​

This module uses the PdfPig library for PDF text extraction. PdfPig is a fully open-source PDF library that:

  • Extracts text content from PDF documents
  • Does not require any external dependencies
  • Works cross-platform

Limitations​

  • Scanned PDFs: Scanned documents that contain images of text (not actual text) will not be extracted correctly. For best results, use PDFs with actual text content.
  • Complex Layouts: Some complex PDF layouts may not preserve exact text formatting.

Supported File Types​

ExtensionMIME Type
.pdfapplication/pdf

Note: The AI Documents feature is provided on demand and is only enabled when another feature that requires it is enabled. To configure document indexing you must enable either the AI Documents - Azure AI Search feature (CrestApps.OrchardCore.AI.Documents.AzureAI) or the AI Documents - Elasticsearch feature (CrestApps.OrchardCore.AI.Documents.Elasticsearch) in Orchard Core admin.