Loading
Loading tool
The tool is loaded only when you open it.
All processing for this tool happens in your browser. Your input is not sent to a server.
About this tool
Read supported document information and XMP fields, edit each store separately or remove supported fields. Reload the saved PDF and preview every actual output page before downloading. Custom metadata is preserved and counted.
Common uses
- Review PDF title, author, creator and date strings before sharing.
- Edit supported Info or XMP values and see which metadata remains outside coverage.
How to use it
- 1.Choose one PDF or load the local example. Unsupported XML shapes, unsafe actions and signed documents reject explicitly.
- 2.Edit Info and XMP fields separately. Blank inputs remove fields. Reset restores the original values; remove supported fields clears only these editors.
- 3.Create the PDF, inspect the saved readback values and all output pages, then save and review with another reader.
Executed local examples
Separate Unicode stores
Local example: Info Title and XMP title = Example 中文 日本語; custom Info = 1, custom XMP = 1. Edit Info Title to Reviewed <tag> 中文 日本語 😀 and XMP title to XMP <tag> 中文 日本語 😀.
Saved stores retain their different Unicode titles; custom Info = 1, custom XMP = 1; 3 pages.
Remove only supported metadata
Load the example, remove supported fields and create the PDF.
Info 8 fields absent; XMP 9 fields absent; custom Info = 1, custom XMP = 1 remain; 3 pages.
Limits and notes
- One PDF / 10 MiB / 50 pages; output 40 MiB. Each value: 1024 UTF-16 units. XMP: compressed and decoded 64 KiB, 2000 XML items, 16 levels. Info: 64 keys. PDF graph: 10000 indirect objects, 50000 visits, 32 direct levels. Reading, worker and preview have cancellable 30-second phase deadlines. Previews: 512px edge, 262144 pixels per page, 16 MiB total; other compressed resources can use more decoder memory.
- Info supports Title, Author, Subject, Keywords, Creator, Producer, CreationDate and ModDate as raw Unicode strings. XMP supports dc:title and dc:description with exactly one x-default rdf:Alt item; dc:creator with one rdf:Seq item; pdf:Keywords, pdf:Producer and xmp:CreatorTool, xmp:CreateDate, xmp:ModifyDate, xmp:MetadataDate as simple strings. Info and XMP are not synchronized. Dates are raw strings, not parsed or corrected. Blank editor inputs remove values; controls, bidi controls and unpaired surrogates reject.
- XMP is UTF-8 XML with one rdf:Description under rdf:RDF, optionally wrapped by x:xmpmeta. Simple supported fields can be attributes. Unfiltered or single FlateDecode streams only; DecodeParms, other filters, DTD/entities, duplicate fields, other languages, multi-author arrays and complex supported shapes reject. Custom properties and attributes are preserved; no XML or document URL is fetched or executed. XML element, attribute and processing-instruction names use the parser’s BMP name subset; supplementary names reject, while text and attribute values support valid Unicode.
- Encrypted PDFs, signatures/permissions, XFA, document/page/additional actions, JavaScript and unsafe actions reject. Safe ordinary URI/GoTo links and unsigned forms may remain. Page geometry, rotation, content/resource/annotation references and unsupported Info/XMP values are checked after save. All actual output pages must preview successfully; no universal losslessness guarantee.
- Removing supported metadata is not full anonymization. Custom Info/XMP, page metadata, attachments, content, annotations, original IDs, hidden/unreachable objects and revision evidence may remain; historical metadata bytes can still occur in the saved file. Attachment detection covers catalog AF and Names/EmbeddedFiles only, not exhaustive absence. File/edit/reset/language/clear/cancel invalidate old output. No upload or input persistence.
Frequently asked questions
Does removing supported fields anonymize the PDF?
No. It removes only the listed reachable Info and XMP fields. Custom metadata and other document structures remain, and older metadata objects or bytes may remain. Review attachments, content and revision evidence separately.
Will editing Info update XMP automatically?
No. They are separate stores and may differ. Edit each field explicitly; the saved receipt reports both stores.
Can I edit dates and multiple authors?
Dates are raw strings; the tool does not validate or normalize calendar values. XMP creator supports exactly one sequence item; multi-author and multilingual alternatives reject rather than losing values.