Support

SlideTouch uses Core ML models that rely on the Neural Engine found in Apple Silicon chips. Intel Macs don't have this hardware, and CPU-only inference would be impractically slow.

SlideTouch is designed specifically for image-based PDFs where text exists as pixels, not as editable text objects. For text-based PDFs, standard PDF editors are more appropriate.

Never. SlideTouch follows a strict non-destructive editing principle. All edits are stored as overlay layers. When you export, a new file is created, leaving the original untouched.

SlideTouch supports 13 languages including English, Korean, Japanese, Chinese (Simplified/Traditional), French, German, Spanish, Portuguese, Russian, Ukrainian, Thai, and Vietnamese. You can select your preferred languages in Settings. Language recognition is powered by the Apple Vision framework.

No. All processing is done entirely on your device. SlideTouch has no network capability and never uploads your files anywhere.

Try using a smaller edit region. The inpainting model works best on small to medium patches. For very large areas, you may need to edit in multiple smaller steps.

SlideTouch loads ML models on demand and unloads them after 30 seconds of inactivity. For very large PDFs (100+ pages), closing pages you're not actively editing will reduce memory usage.