Title
A short display title written as Dublin Core XMP.
Private, local XMP editing
Image metadata editor
Inspect an image, edit a small set of portable metadata fields, and download a newly re-encoded copy. The browser re-scans every written field and never overwrites your original.
Local image metadata editor
Write a fresh, verified metadata set
Existing metadata is inspected first. The new copy is re-encoded, receives only the fields you enter below, and is scanned again before download.
Drop one image here
Choose a JPEG, PNG, or static WebP. The file stays on this device while the browser reads, rewrites, and verifies its metadata.
JPEG, PNG, static WebPOne file, up to 25 MB
On-device
No processing upload
New copy
Original stays unchanged
Re-scanned
Written fields verified
Batch editor
Apply verified XMP fields to several images
Use one local preset for up to five JPEG, PNG, or static WebP files. Each new copy is re-scanned before it becomes downloadable.
Add batch images
Up to 5 files · 25 MB each · original files remain unchanged
Editable XMP fields
Five portable fields, with a narrow promise
The editor deliberately avoids GPS, device serial numbers, AI workflow parameters, timestamps, and provenance claims. Those fields need different intent and stronger validation.
Description
A caption or context note stored in a portable XMP description field.
Creator
A person or organization name chosen by the user; authorship is not verified.
Copyright
A rights notice written exactly as entered, without validating ownership.
Keywords
A small Dublin Core subject list for cataloging and retrieval.
When to edit metadata instead of removing it
Removing metadata is the right default for private sharing, but some files should carry accurate information: who made the image, who holds the rights, and what it shows. This editor covers those cases by discarding the old record and writing five portable fields that you control.
Credit your work with the Creator field
Photographers, illustrators, and studios often publish images that have lost their byline through resizing, exports, or platform re-encoding. Writing your name or studio into the Creator field gives downstream viewers a portable, machine-readable attribution that many asset managers and operating systems can read. The field records exactly what you type. The tool does not verify authorship, so use the name or imprint you actually publish under.
State licensing terms in the Copyright field
Use the Copyright field for a rights notice and a way to reach you, for example a © line with a year, a license name, or a contact address. A clear notice makes honest reuse easier and misuse harder to excuse. The tool writes the statement exactly as entered and does not verify ownership, licensing authority, or identity; the accuracy of the claim remains your responsibility.
Prepare files for clients and stock delivery
Client handoffs and stock submissions go smoother when every file arrives with a consistent title, a usable caption, and searchable keywords. Instead of relying on filenames, write those values into the image itself so they survive renaming and travel with the copy. Because the editor verifies each written field by reading it back, you know the delivered file actually carries the description you intended.
Editing here doubles as a privacy cleanup
The editor never patches metadata into the existing file. It re-encodes the visible pixels into a new copy, discards the original metadata containers, including GPS, device details, timestamps, and AI workflow fields, and then writes only the five fields you filled in. The result is one predictable step: supported personal-risk metadata is left behind while your attribution goes in, and the output is re-scanned to confirm both.
Inspect → write → verify
Editing remains an evidence-first workflow
The tool does not show a success toast immediately after writing bytes. It checks the new file and only enables download after the requested fields are found again.
- 01
Inspect the original
Validate the real file signature, apply the same safety limits, and load supported existing values into the form.
- 02
Create a fresh copy
Re-encode visible pixels, discard the original metadata container, and write only the chosen XMP fields.
- 03
Re-scan every field
Read the output with the same scanner, confirm the requested values, and provide a share-safe report without those values.
Why the editor writes Dublin Core XMP
Of all the places image metadata can live, a Dublin Core XMP packet is the most portable home for titles, descriptions, creators, rights, and keywords. The editor writes exactly one fresh packet per copy, then proves every field landed by reading the output back.
An open standard most software can read
Dublin Core inside XMP is a documented, openly specified vocabulary rather than a vendor-proprietary block. Mainstream photo managers, editing suites, and the file-info panels of major operating systems can generally read these fields, which makes them a sensible default for attribution and rights notices. No standard guarantees display everywhere, though: each application chooses which fields to surface, so a written value can be present yet not shown in a given program.
A fresh packet, not an in-place patch
In-place editors modify bytes inside an existing container, which can leave stale duplicates, orphaned thumbnails, or conflicting values behind. This editor takes the predictable path instead: the old metadata is discarded with the re-encode, and a single new XMP packet is written containing only your five fields. Because the output's entire supported metadata surface is known, the re-scan can check it meaningfully. There is no leftover state to guess about.
Every field is read back before download
Writing bytes is not the same as proving they arrived. After creating the new copy, the editor re-opens it with the same scanner used for inspection and compares every written field against what you entered, expecting an exact match. Only when each requested value is found again does the download unlock. If verification fails, the tool says so plainly instead of handing you an unchecked file.
Important limits
This editor favors a predictable privacy boundary
It is not a lossless EXIF patcher or a provenance authoring system. The behavior below is part of the product, not fine print.
The image is re-encoded
Compression, file size, color profiles, and subtle pixel values may change even when dimensions and visible content remain the same.
Old metadata is not preserved
GPS, device, timestamps, AI workflow, thumbnails, and unsupported proprietary fields are not copied to the edited output.
No C2PA authoring
Possible markers require explicit removal consent. The tool cannot validate a signature or create replacement Content Credentials.
Images only
The V1 editor supports JPEG, PNG, and static WebP. PDF, HEIC, RAW, animation, audio, video, and Office documents remain outside this page.
Metadata editor questions
Know what will change before creating a copy
These answers separate portable XMP editing from EXIF patching, provenance authoring, and lossless container changes.
Does the image leave my device?
No. Inspection, re-encoding, XMP writing, verification, and download run in your browser. The tool does not send the image, filename, metadata values, or file hash to a processing API.
Which metadata fields can I edit?
The V1 editor writes title, description, creator, copyright, and keywords in a fresh Dublin Core XMP packet. Empty fields are omitted.
Does it edit every EXIF field?
No. GPS, camera details, serial numbers, dates, orientation, thumbnails, MakerNotes, AI workflow parameters, and proprietary fields are not editable here. The old metadata is discarded rather than patched in place.
Will image quality or file size change?
Yes, they can. The browser re-encodes visible pixels before writing the fresh XMP packet. The report shows the output format, dimensions, and input-to-output size, but file size is not a metadata verification signal.
Can it create or preserve Content Credentials?
No. C2PA support is conservative marker detection only. If a marker is found, creating an edited copy requires explicit consent because re-encoding removes or detaches it; the tool does not validate or forge credentials.
Which formats are supported?
One JPEG, PNG, or static WebP up to 25 MB. Animated WebP, APNG, multi-asset JPEG, HEIC, RAW, malformed files, and formats without the tested browser encoder path are rejected.
Can I edit GPS, capture date, or camera fields?
No, and that is deliberate. The editor writes exactly five Dublin Core fields: title, description, creator, copyright, and keywords. Everything else from the original, including GPS coordinates, capture dates, camera and lens details, serial numbers, and AI workflow parameters, is discarded rather than carried over or edited. Keeping the writable surface this small is a privacy boundary: you cannot accidentally preserve a location trail, and the output never contains fields the tool could not verify. If you need full EXIF editing, this is intentionally not that tool.
Will Lightroom, Photoshop, or my operating system read the fields?
Dublin Core XMP is widely supported, so applications like Lightroom and Photoshop, most digital asset managers, and the file-information panels of major operating systems can generally read title, description, creator, copyright, and keywords. What each program displays differs, however: some show every field, some surface only a subset, and a few label them differently. The tool can verify that the fields exist in the file; it cannot guarantee that every application will show all of them.
Does writing a copyright notice prove I own the image?
No. The Copyright and Creator fields record a statement you chose to make; the tool does not check your identity, your rights, or any license, and an embedded notice is not a registration or legal proof of ownership. Metadata fields can also be rewritten by anyone with a copy of the file. Treat the field as a clear public declaration and a contact point. It is valuable for honest reuse, but not evidence on its own.
Only need to inspect?
Open the read-only Metadata Viewer to see supported fields without re-encoding or writing a new file.
Open Metadata ViewerWant a metadata-cleaned copy?
Use AI Metadata Remover when you want to discard supported file-level metadata without adding a fresh authoring set.
Open AI Metadata Remover