# Changelog of PDF to Markdown & JSON for RAG (OCR, DOCX, PPTX, XLSX) (`maxatron/docmd`) Actor

- **URL**: https://apify.com/maxatron/docmd/changelog.md
- **Full Actor documentation**: https://apify.com/maxatron/docmd.md

## Changelog

### 0.2 (2026-09-26)

- **New fast mode (default).** Reads the PDF's text layer directly and extracts tables with a
  line-based detector, with OCR only on scanned pages. A normal page converts in about half a second.
- **Accurate mode** keeps full layout analysis and table-structure recognition for complex documents.
  It needs 4096 MB of memory, and a run with less stops before charging anything.
- The file type is now detected from the downloaded contents, so URLs without an extension work.
- Downloads are capped at 100 MB. Failed, oversize or unsupported documents are never charged.
- Runs respect your maximum-cost setting and stop cleanly before exceeding it.
- Output items now include the `mode` used.

### 0.1 (2026-09-25)

- Initial private version.