Verify metadata removed: compare a file before and after
Verify metadata removed: compare a file before and after?
Drop the original and the cleaned copy in this tab to compare JPEG, PNG, WebP, MP4, MOV, PDF Info and catalog XMP, and Office property parts, then read which blocks and fields remain, along with both SHA-256 hashes, without uploading either file.
Last updated
Drop the original and the cleaned copy in this tab to compare JPEG, PNG, WebP, MP4, MOV, PDF Info and catalog XMP, and Office property parts, then read which blocks and fields remain, along with both SHA-256 hashes, without uploading either file.
Tool added 2026-10-18
Compare two files
No file chosen.
No file chosen.
Choose a before file and an after file to compare.
How to read the result
Each row is one block: a JPEG segment, a PNG chunk, a WebP RIFF chunk, or an MP4 box. “Removed” means the block is in the before file and not in the after file. “Zeroed” and the padding row appear for MP4, where the cleaner overwrites a box with empty free padding so the file keeps its length. A color profile is listed apart from metadata because it describes color, not a person.
The field lists come from the same parsers as the Image Metadata Viewer and the Video Metadata Viewer. If a field is not in the list, the page could not read it, which is not proof that it is absent.
Two things it cannot see: pixels and the file system. A date stamp drawn into the picture stays, and so does the file name. See whether a screenshot removes metadata.
Measured examples
On a GPS-tagged JPEG the Metadata Cleaner removed the EXIF segment and the GPS position and left an ICC profile: JPG details. A PNG lost its text chunk: PNG details. A WebP lost its EXIF chunk and kept a color profile: WebP details. An MP4 kept its creation time after cleaning: MP4 details.
Browser-side proof that nothing is sent: the methodology page explains how to watch the Network tab. To clean a file first, use the Metadata Cleaner, the Office property tool, or the PDF Info and XMP remover. What the second read can and cannot prove is in the verify-after-strip note.
FAQ
Are my files uploaded?
No. The page reads each file with the browser's File API, parses it in this tab, and hashes it with Web Crypto. The only network requests are the page's own scripts. You can confirm in the Network tab of your browser's developer tools (F12) while you add a file.
What does the page compare?
The blocks each file carries (JPEG segments, PNG chunks, WebP RIFF chunks, MP4 and MOV boxes, PDF Info and catalog XMP, and Office core, app, and custom parts) and the fields the site's readers can parse from them, such as GPS, camera, software, dates, author, and document properties.
What does a clean verdict mean?
Only that the after file has no metadata blocks or readable fields this page knows how to find. It does not look at pixels, the file name, or file dates on disk. PDF Info and catalog XMP, and Office core, app, and custom parts, are included when the file parses. HEIC is not.
Why does an MP4 still show a creation time?
Some cleaners, including this site's, neutralize user-data boxes and leave the movie header, which holds creation and modification times. The page flags that as remaining metadata.
What is the size limit?
512 MB per file, because each file is read into memory.