# Changelog of PDF to Text & Markdown with OCR for Scans (19 Languages) (`prime619/pdf-to-text-markdown-ocr`) Actor

- **URL**: https://apify.com/prime619/pdf-to-text-markdown-ocr/changelog.md
- **Full Actor documentation**: https://apify.com/prime619/pdf-to-text-markdown-ocr.md

## Changelog

### 0.1.7 (2026-10-09)

- Text pages beyond the 15 charged pages of a PDF use a fast plain-text mode (no Markdown structure); new field `fastTextPages`. 88-page PDF: ~4× faster.

### 0.1.5 (2026-10-09)

- Scan cleanup before OCR: lighting correction for phone photos, deskew (±8°), sideways/upside-down pages turned upright, adaptive thresholding retry, low-confidence noise lines dropped; per-page `scanFixes`.
- German Fraktur model, auto-detected; safer automatic language detection on poor pages.
- Text pages are charged for at most 15 pages per PDF (`chargedEvents` in each result).
- Tested on 36 real public-domain scans in 6 languages (results in README).

### 0.1.0 (2026-10-09)

- First private version (not published): text layer + Tesseract OCR (19 languages, auto-detect), Markdown with headings/lists/tables, RAG chunks, form fields, only-new-files.
- Fixes before first measurement: key-value store schema accepted only text/markdown (runs failed when saving .txt); a file that the cost limit stopped before its first page is no longer charged; rotated side text is skipped; word spacing for tightly set PDFs (LaTeX).