Everything you need for PDFs

Which PDF tool do you need?

Most PDF problems are one of five, and the right tool is usually decided by what you want to be true of the file afterwards rather than by what is wrong with it now.

The file is too big to email or upload

Start with Compress PDF, but only if the bulk is scans or photographs. It rasterizes every page to JPEG, so text stops being selectable and searchable — a real cost, and the reason the tool rejects a result that came out no smaller than the original. If the document is mostly text, compression will not help much; removing pages with Split PDF usually will.

You need part of a document, not all of it

Split PDF in range mode builds one new PDF from the pages you pick. Every-page mode gives you a separate single-page PDF for each page instead. The original is never modified either way, and overlapping selections are rejected rather than silently duplicating pages.

The pages are in the wrong order, or the wrong way up

Rearrange PDF Pages changes only the sequence — nothing is cropped or resized. Rotate PDF turns pages in 90-degree steps by setting the page rotation flag, so it does not re-render your text as an image. Neither tool touches the content of a page.

Several files should be one file

Merge PDF appends complete documents in the order you arrange them, preserving each source page size. Pages are not resized to match one another, so a mixed A4-and-Letter merge stays mixed.

You need the pages as images, or images as a PDF

Going out: PDF to JPG for scans and photographic pages, PDF to PNG for text, line art and screenshots. PNG is lossless and larger, JPG is lossy and smaller, and choosing by what is on the page matters more than choosing by preference. Going in: JPG to PDF and PNG to PDF.

The document should be marked before you send it

Watermark PDF stamps text or a logo across the pages you choose, with control over opacity, rotation, position and tiling. Page size, page order and page count are unchanged.

What these tools will not do

Every tool here builds a new document rather than editing the original in place, and that one fact explains most of the limits worth knowing before you start.

• Interactive features may not survive. Bookmarks, form fields, annotations, internal links, metadata and digital signatures belong to the document that had them, and a newly assembled document does not inherit them reliably. Each tool page states its own risk; if any of these matter, keep the original.
• Encrypted files are rejected, not unlocked. A password-protected PDF is refused rather than bypassed. That is a deliberate boundary, not a gap.
• Rasterizing is one-way. Compress PDF and DOCX to PDF both produce pages made of pixels. You cannot select the text, search it, or recover the vectors afterwards, so keep the source file.
• A watermark is not access control. The mark is drawn into the page content, which makes it awkward to remove but does not stop anyone copying, printing or redistributing the file.
• Your device sets the ceiling. Because the work happens in your browser, the document and the output both have to fit in its memory. The size limits on each tool reduce accidents; they do not guarantee that every allowed file will finish on every phone.
• Nothing here reads text from a scan. No OCR runs, so a scanned page stays a picture of words in every output format.

The wider account of where processing happens, what stays in memory, and how to check any of it for yourself is on Security and File Processing.