Every tool has edges. Knowing them before you start is worth more than discovering them halfway through.
What this tool will not do
Password-protected and encrypted PDFs are reported, not opened. This tool never removes an access control.
There is no compression feature, because the usual browser technique destroys selectable text and often makes text documents larger.
There is no OCR, so text inside a scanned image stays an image and cannot be searched or selected.
PDFs above 50 MB and images above 30 MB are refused, because your device does the processing.
Text watermarks use the built-in PDF fonts, which cover Latin characters only. Non-Latin text is reported rather than silently dropped; use an image watermark instead.
Only the first 60 page thumbnails are rendered in the reorder view. Later pages are still included in the output but are not previewed.
Digital signatures are not preserved. Any edit invalidates an existing signature, as it would in any PDF editor.
Interactive form fields and annotations are carried across where the format allows, but complex forms may not survive reordering.
Watermarks are visible marks only. They are not encryption, access control, or copy protection.
Text watermarks support Latin characters only, because they use the fonts built into the PDF format.
The PDF format embeds PNG and JPEG only. Other formats are converted by your browser first, and HEIC conversion needs operating-system support.
Images larger than the device memory budget are automatically reduced before embedding.
Converting PDF pages to images loses selectable text: the result is a picture of the text and cannot be searched or read by a screen reader.
Rendering many pages at high resolution is memory-intensive and may be slow on older phones.
Watermarks are drawn on top of existing page content and cannot be removed by this tool afterwards.