Remove DALL·E Metadata (C2PA)
Drop images here or click to upload
JPG, PNG, WebP, AVIF, HEIC · Max 15MB · Batch up to 30 · Not GIF
OpenAI attaches C2PA-style provenance to many downloads.
Recommended workflow: metadata checker → remover
Each AI exporter leaves different fingerprints in your file. Inspect read-only before you strip fields — then use the remover on this page for the upload you will publish.
- Step 1 — inspect your file: AI metadata checker.
- Upload your export to the AI metadata checker (read-only, browser-only)
- Review C2PA, XMP, PNG chunks, and EXIF software strings
- Use the remover above with matching options enabled
- Download the cleaned JPG, PNG, or WebP
- Upload to Instagram, Facebook, Pinterest, TikTok, or LinkedIn
After Photoshop, Canva, or compositing
A raw AI download is not automatically upload-ready. Generative Fill, Canva AI backgrounds, or compositing can add C2PA or IPTC AI fields on top of the original export. Always clean the final flat raster you publish — not intermediate layers or PSD masters.
Why DALL·E files often show Meta's AI Info label
Meta has publicly described C2PA as an important signal for AI disclosure on photos. OpenAI's image pipeline writes content credentials into many downloads so downstream apps can reason about provenance.
That supports transparency workflows — and it is what you typically strip when your goal is an upload without metadata-driven disclosure, as long as you still meet any legal duties where you operate.
What you should remove for OpenAI exports
If you composite a DALL·E plate into Photoshop, you may inherit Adobe C2PA as well — flatten and process the final export once.
- C2PA / JUMBF manifests and embedded assertion stores
- Auxiliary XMP generation or software fields if present
- Obvious default filenames after you have cleaned bytes
Clean this file before you upload
Strip C2PA and XMP in your browser, then publish the cleaned copy.
DALL·E vs Midjourney vs Stable Diffusion
Each toolchain fingerprints differently:
- OpenAI → C2PA-first, Midjourney → XMP-heavy,
- ComfyUI / SD → PNG text chunks.
Compare stacks in depth
Workflow
- Download your DALL·E or Labs image.
- Upload here with C2PA removal enabled.
- Download the cleaned PNG or JPEG.
- Rename if the default name advertises the generator.
- Optional: confirm empty readout in the metadata checker.
- Post to Instagram, Facebook, Pinterest, or TikTok.
What this tool removes (OpenAI exports)
- C2PA / JUMBF content credentials
- XMP AI generation parameters
- PNG text chunks when present
- IPTC AI attribution
- EXIF software strings
Does not remove pixel watermarks. See disclaimer .
FAQ
Do DALL·E and OpenAI Labs images include C2PA?
Many DALL·E, Labs, and API exports ship with C2PA manifests and related JUMBF boxes inside PNG or JPEG containers. Meta has described C2PA as a key signal for Instagram and Facebook; when that metadata is present, uploads often show an "AI Info"-style disclosure shortly after upload. For native ChatGPT image downloads, see our ChatGPT Image page.
Will filenames like DALL-E-2025.png trigger labels alone?
Filenames can be weak auxiliary signals on some platforms. Rename after you strip C2PA so you are not stacking easy heuristics on top of cleared bytes.
Does removing C2PA change pixels?
No. C2PA data is stored separately from the decoded bitmap. Removing it does not alter visible image content.
API responses behave the same as ChatGPT UI downloads?
Treat every binary response as untrusted: some code paths strip metadata, others preserve it. Always verify the bytes you ship.
Does Pinterest read C2PA the same way?
Pinterest publicly discusses layered GenAI detection; C2PA is still a high-signal block on many assets. Clearing C2PA is the correct first step before pinning OpenAI exports — see our Pinterest guide.
Try it now
Confirm C2PA is gone with our AI metadata checker.