Skip to main content

Document Conversion

Document Conversion — Docling​

Model NameModel IDDescription
DocumentdoclingConverts documents to Markdown for AI consumption

Supported formats​

PDF, DOCX, PPTX, XLSX, HTML

How it works​

  1. The document is parsed and converted to Markdown.
  2. Images and diagrams are described as text by "Qwen3.5-397B-A17B" LLM.
  3. The resulting Markdown can be injected as LLM context (CAG / Cached RAG / Vision RAG).
note

All tokens consumed during image-to-text conversion are charged under the Qwen3.5-397B-A17B model.