Metadata is the information a file carries about itself: GPS coordinates and serial numbers in photos, recording dates and sensor tracks in videos, author names and tracked changes in Word, Excel and PowerPoint. This page takes all of those in one batch, removes the fields without re-encoding any picture or sound, and rereads every output to list the kinds of metadata it still finds.
Why one page for every kind of file
Most people do not share one file at a time. You send a folder to a contractor, attach five photos and a spreadsheet to a listing, or hand a client a video and the document that explains it. Each of those files was written by a different program, and each program leaves its own trail.
A phone photo can hold the spot where you stood, the exact second you pressed the shutter and a code that identifies the phone. A clip from an action camera can carry a second-by-second GPS track. A Word document can still contain the name of every person who edited it, plus text they deleted with tracking switched on. None of this shows up when you open the file normally.
This page sorts a mixed batch for you. Photos, videos and Office documents each go through the routine written for their format, so you do not have to visit three tools to prepare one email.
How to clean a mixed batch
- Set the options first: under "What to remove", keep Everything for photos or pick Location only; tick Keep subtitle tracks only if your videos need captions; choose how Word's tracked changes should be handled.
- Press Choose files or drop files on the box. You can mix JPEG, PNG, WebP, HEIC, MP4, MOV, DOCX, XLSX and PPTX in one go. Processing starts as soon as the files arrive.
- Read each result card. It lists what was found in the file, what was removed and what the tool still finds in the cleaned copy.
- Save each file with its Download button, or use Download all (zip) when the batch has more than one file. On a phone you may also see a Share button.
- If you change an option after a run, press Start to process the same files again with the new setting.
What happens to each kind of file
| File type | Removed | Kept on purpose |
|---|---|---|
| Photos (JPEG, PNG, WebP, HEIC) | EXIF camera data, GPS, maker notes, XMP, IPTC, Content Credentials, comments, the preview image inside EXIF, PNG text notes, extra images after a JPEG's end marker | The rotation flag and the color profile (with the short description text a profile holds), so the picture looks the same |
| Videos (MP4, MOV, plus .LRF and .LRV proxies) | Data tracks such as GPS telemetry and timecode, user-data and metadata blocks, vendor boxes, creation and modification times, track names | Picture and sound, byte for byte |
| DJI .SRT subtitles | Coordinates, home point, altitude and distance fields | Line numbers, timing, date and time, exposure settings |
| Word, Excel, PowerPoint | Names of the author and company, comments with the people who wrote them, tracked-change names, editing-session ids, hidden text, custom properties, the folder part of a file path in links and wherever a path stands alone, previews, photo metadata inside the document | The document's visible content; PowerPoint speaker notes; digital signatures, data connections and add-in data, which the result card names |
Each row has a page that explains the details. For photos, read remove metadata from a photo. For video, see remove metadata from video. The Office pages cover Word, Excel and PowerPoint separately because each program hides different things.
Options that change the result
Everything or Location only. The default removes every photo field in the table above. Location only clears GPS, the blocks that can repeat a place name and EXIF entries the cleaner cannot put a name to, but leaves the date and camera model. Use it when the date matters to the person receiving the file. See what Location only keeps for the trade-offs.
Keep subtitle tracks. Off by default: certain drone models put their position into a subtitle track inside the video. Tick it only for videos whose captions you wrote yourself, and check them afterwards.
Tracked changes in Word files. Accept all changes resolves the edits and leaves the final text, with the reviewer names taken out along with the history. With Keep changes, hide who made them, the edits stay in place for the reader but show as made by "Author", stripped of the dates they were made on (moved text gets the fixed date 1 January 2000, because the format wants one there). The first choice alters the wording of the document, so read the result before sending it.
Formats this page does not handle
- PDF. Use a separate tool to edit or clear PDF document properties.
- Old Office formats (.doc, .xls, .ppt). Open them in Office and save as .docx, .xlsx or .pptx first.
- Insta360 .insv. GPS and the serial number sit in a block at the end of the file that this tool does not edit yet. Export an MP4 from the Insta360 app, then clean the MP4.
- Camera RAW, GIF, MP3, WAV or FLAC audio, and archives are not supported. (An .m4a file is an MP4 container inside, so it passes through the video cleaner.)
- Unusual photo files. JPEGs with arithmetic coding or in hierarchical mode, HEIC image sequences and animated AVIF are turned away; re-save the picture as a regular JPEG first.
A file that cannot be read safely is left untouched, and the card says why. The other files in the batch carry on.
What these cleaners cannot do
A card that lists nothing as left reports only on what the tool checks for; the file may still hold a clue about you. In short:
- what a picture or a video shows, and what a recording says, is never examined;
- Location only keeps the camera, software and lens fields and the owner, artist and copyright names;
- color profiles stay and are listed, and so is the separate preview inside a HEIC;
- in a video, the encoder name field is cleared in H.264, H.265 and ProRes but stays in other formats, and a vendor code or an encoder's banner can stay in the codec settings or ahead of the first frame; the result lists what stays;
- in Office files, a few default names are not searched for, a path that a formula joins together is not recognized, and a folder path in a rare spelling inside a longer text can be missed.
Each of these is spelled out, with examples, in the limits section of the methodology page.
Where your files go
They stay on your device. The page reads each file from your disk into the browser tab, cleans a copy in memory and hands that copy back as a download. No file, file name or fragment of content is sent to a server. Closing the tab discards everything the page was holding.
The zip for a batch is built in the same tab. Every entry gets the fixed date 1 January 1980, so the archive does not record when you cleaned the files.
How the cleaning works
Each format has its own parser. Photos are edited at the byte level: metadata blocks are cut out and the compressed image data is copied unchanged, so nothing is decoded or redrawn. Videos are patched in place: unwanted tracks and boxes are overwritten with zeros and relabeled as free space, which keeps every offset valid and the file size identical. Office files are zip packages of XML; the tool rewrites the parts that hold names and history, then a separate check reads the finished file once more and reports what it cannot vouch for. All of this runs in background workers, so the page stays responsive during a large batch.
To confirm for yourself that no upload takes place, follow the Network panel steps on the methodology page while you clean a file.
FAQ
Can I really mix photos, videos and documents in one batch?
Yes. The page checks each file's type and sends it to the matching cleaner. A batch of 30 photos, two MOV clips and a spreadsheet is fine. Results appear one card per file, in the order the files were processed.
Is there a limit on how many files I can add?
There is no fixed cap. Batches of 50 photos have been tested. Very large videos are handled in slices, so their size matters less than you might expect, but a phone with little free memory may struggle with dozens of big files at once.
Will my photos lose quality?
No. The compressed image data is copied exactly. The cleaned photo decodes to the same pixels as the original, and the file usually gets a little smaller because metadata bytes are gone.
Why does the cleaned file have "-clean" in its name?
So you can tell it apart from the original. The original file on your device is never changed. Keep in mind that the name itself can reveal a date or a device; rename the copy if that matters.
Does removing metadata make a file anonymous?
No. It removes what the file says about itself. The picture can still show a street sign, a face or a reflection, and a document's text can name people. Read the content before you share it.
What does "Left in the file" mean on a result card?
After cleaning, the tool reads the output again and lists each kind of metadata it can still find. For photos that is usually the color profile, plus the rotation flag when the picture is saved on its side; each is left in deliberately. A HEIC that carries a separate thumbnail lists it too, as an embedded preview image. For an Office file the list moves into a box that opens the card, headed Still in the cleaned copy, and each item shows which part of the file it is in. Anything listed is worth a look before sharing.
Sources
Checked 2026-10-11.
- ExifTool tag names: EXIF
- ExifTool tag names: QuickTime
- Microsoft's guide to the Document Inspector
- GoPro GPMF parser documentation
Last updated: