What changed
Scanned documents are where many “chat with your files” workflows still break.
Microsoft released MarkItDown 0.1.8 with dozens of small fixes and says typical inputs should behave largely like version 0.1.7.
How it works
The release includes a MarkItDown OCR plugin path, extending a tool that converts common document formats into Markdown for downstream processing.
Markdown gives agents and language models a simpler text structure than many original office-file formats, but OCR quality still depends on the scan.
Why it matters
The useful test is a messy real document: compare headings, tables and numbers against the source before trusting an agent with the converted text.
The source link below contains the technical details and release context; the result should be judged on the specific workflow it improves, not the announcement alone.
Source published 2026-09-22. Coverage is based on the maker’s announcement and demonstration.