PDF Workbench does the page-level jobs: combine several PDFs into one, pull a range out into its own file, turn pages that were scanned sideways, and drag pages into a new order. It also shows the document metadata, the author, the producer, the creation date, and lets you clear it.
The PDF is opened with the browser File API and rewritten on your machine. Nothing is uploaded, and the output has no watermark.
No limit beyond your device memory, and no watermark. The output is a plain PDF.
Page content, text, and images are kept. Interactive form fields and the outline tree may not survive a merge; check the result if your source has them.
Stripping metadata and dropping pages helps a little. Heavy image compression is the job of Image Squeeze, not this tool.
No. The file never leaves the tab. Close it and nothing is retained.