Scanned contracts, photographed pages, and image-only PDFs now convert to searchable Markdown: the converter detects pages without a text layer, runs OCR, and merges the recognized text into the standard output.
The preview panel now builds a nested, clickable outline from the headings of every converted file, so long documents can be inspected section by section instead of scroll by scroll.
Links inside PDF files now convert to standard Markdown links, keeping anchor text and URLs intact for external, in-document, and email targets.
DOCX files with footnotes and endnotes now convert to Markdown with note markers kept inline and note bodies collected as reference-style links.
EPUB files now convert to Markdown with chapter titles promoted to headings, inline formatting preserved, and footnotes collected as reference links.
Plain-text comma- and tab-separated files now convert to properly aligned Markdown pipe tables, with header detection, quoted-field parsing, and empty-cell handling.
Extract tables from PDFs, DOCX, and images in batch while preserving merged cells, spanning headers, and formatting.
Converted files now start with a YAML front matter block recording the source file name, format, page or sheet count, and conversion timestamp.
Transcribed audio files now export Markdown with paragraph-level timestamps.
Excel formulas are now exported as inline code in the Markdown output.
Excel formulas are now exported as inline code in the Markdown output, so downstream readers can trace each cell back to its calculation.
Footnotes and endnotes in Word documents now convert to Markdown reference-style links with stable anchors instead of being dropped.
Complex HTML tables with merged cells now convert to Markdown with accurate pipe-table representations instead of losing cell relationships.
Convert multiple files simultaneously with parallel workers, reducing total batch time by up to 3x for large uploads.
MarkItDown Online now flags files with broken heading hierarchies or empty sections before the ZIP export completes.
See how long each file took to convert so slow outliers are obvious before you download the ZIP.
Scan H1–H3 structure across a converted batch so broken outlines are caught before ZIP download.
Convert a folder of files and download everything as one ZIP instead of clicking through individual downloads.
Slide decks converted through MarkItDown Online now include speaker notes in a dedicated section below each slide's main content.
Workbook conversion now does a better job keeping multi-row headers and merged-cell tables readable in Markdown.
Spot-check a random sample from each batch to catch systematic conversion issues before they compound across hundreds of files.
Convert PDFs to clean Markdown before chunking so retrieval and prompting stay structured.
Why automated PDF/DOCX/PPTX-to-Markdown conversion beats manual cleanup when preparing documents for LLM pipelines.
Download all converted files as a single ZIP archive and compare outputs side-by-side in a split-pane preview.
Recent updates focus on more reliable extraction for Office files, HTML exports, and image-heavy reports, so Markdown output stays readable with less cleanup.
The upload and preview pipeline is tuned for lower waiting time and clearer feedback, making it easier to inspect output before downloading.
The next wave of work targets multilingual UX, more polished legal pages, and a cleaner content hub around docs and tutorials.
