✦ Built on Microsoft MarkItDown
Turn anything into
Markdown.
Drop your files. Paste a URL. Walk away with clean, structured Markdown — no terminal required.
Drop your files. Paste a URL. Walk away with clean, structured Markdown — no terminal required.
Drag files directly into the queue, add entire folders (top-level), or paste website URLs. Batch processing handles dozens of documents in one session.
Elapsed timer, per-document page counter, and a visual progress bar keep you informed. Pause, resume, or cancel at any time — including in-flight HTTP requests.
Toggle between rendered Markdown and raw source. Select individual files from results, copy output, or export as one combined file or separate files.
Paste any web URL and convert article content to Markdown via the Defuddle API. Responses include YAML frontmatter metadata when available.
Three OCR paths to handle any document: Azure Document Intelligence, local Tesseract, or OpenAI-compatible vision endpoints like LM Studio.
Four theme modes including the custom Perfect Dark palette with refined nav and title chrome, plus a centered content column on ultrawide displays.
Elapsed time, document count, and per-document page counter update in real time during conversion.
Point the app at any OpenAI-style vision endpoint for rasterized PDFs and images. Optional "always OCR" toggle for stubborn scans.
Cancel tears down in-flight HTTP requests where applicable, with clearer progress feedback across PDF and model operations.
Import top-level files from a directory with filter support. English and Chinese strings cover every new UI element.
uv sync
uv run python -m markitdowngui.mainpip install -e .[dev]
python -m markitdowngui.mainuv pip install -e .[dev]
pyinstaller MarkItDown.spec --clean --noconfirmCloud-powered OCR with 500 free pages/month. Set your endpoint in Settings and the AZURE_OCR_API_KEY env var.
Fully offline. Install the Tesseract binary and set its path in Settings if it's not on your system PATH.
Point at LM Studio or any OpenAI-style endpoint. Configure base URL, model, and prompt directly in-app. Optional "always OCR" for difficult PDFs.
| Shortcut | Action |
|---|---|
| Ctrl + O | Open files |
| Ctrl + S | Save output |
| Ctrl + C | Copy output |
| Ctrl + B | Start conversion |
| Ctrl + P | Pause / Resume |
| Ctrl + L | Clear queue |
| Ctrl + K | Show shortcuts |
| Esc | Cancel conversion |