What's New in 3.3: • Layout reconstruction now matches the original PDF more closely — text, fonts, colors, and reading order stay true to the source. • Improved image preprocessing — shadows, scanner noise, and low-contrast pages are handled before OCR runs. • Faster conversion engine — a typical deck finishes in under 10 seconds. • Better mixed-language OCR — Chinese, Japanese, Korean, and Latin scripts in the same document no longer confuse each other.

Smarter layouts. Sharper text frames. • Multi-column reading order — AI now reads top-to-bottom, left-to-right, so dense NotebookLM slides keep their original order. • Pro text frames — every line becomes a real editable box with its original font, size, and alignment. • Pixel-perfect font matching — original font family, weight, and size preserved automatically. • Cleaner layers — text and background graphics are now cleanly separated, no more overlap. • Faster on complex pages — new processing pipeline cuts recognition time by up to 40%.

Better conversion quality across the board. We rebuilt the conversion engine with a new 5-stage pipeline: Render → Layout Detect → OCR → Style Inference → Emit. Each stage now uses typed contracts and isolated error handling, making the engine faster, more accurate, and more reliable on long decks and dense layouts. What you'll notice: • Cleaner text recognition, especially on smaller fonts • Tighter alignment between text boxes and the original slide • Faster conversion on multi-page decks • More stable handling of complex multi-column layouts If something looks off, please send feedback from Settings → Feedback — we read every note.

SlideEdit comes to Mac. And everything else gets better. • Now on Mac. Keyboard shortcuts, Magic Mouse, external display, and full multitasking. • Real-time text editing — fix a typo, change a number, rewrite a bullet, right inside the slide. • Smarter OCR — better with handwriting, low-quality scans, and multi-column layouts. • 25 languages. Including yours. No upload. No account. No cloud. Privacy first, like always.

What's new in 2.1 — a major engine upgrade. NEW: Import more formats • Open Word (.docx), Excel (.xlsx), PowerPoint (.pptx), plain text, and RTF • Save any of them as a beautifully rendered PDF • Vector text, smart layout, true fonts — straight to your Files app NEW: Real-time text editing • Edit any text box right after conversion • Change words, font size, color, alignment, and position • Fine-tune slide by slide before exporting IMPROVED: Layout precision • Smarter reading flow for multi-column research papers • Better table and chart detection • Stronger font recognition with our on-device AI model IMPROVED: OCR • Auto-detect Simplified Chinese, Traditional Chinese, English, Japanese, Korean, French, German, Spanish, Portuguese, Italian, Russian, and more • Faster accurate recognition on iPhone and iPad As always, your files stay on your device. No account, no upload, no cloud.

Improve layout precision

We've integrated advanced AI technology to significantly improve font recognition accuracy and layout precision. Your documents now feature smarter text rendering with optimized typography, delivering a more professional and polished presentation experience.

Added real-time editing for extracted text Improved OCR accuracy and layout reconstruction Enhanced PPTX export for better editability

Experience crystal-clear text! New version boosts typography precision by up to 10%.