IMG Performer vs. IP2C
IP2C detects AI images through metadata and optical analysis and burns a label into the image. IMG Performer detects locally, never alters the original, and bundles image optimization and alt text.
Short version: IP2C detects AI images through metadata and optical analysis and burns a label directly into the image — a solid labeling tool. IMG Performer does the same detection locally, without sending images to an external provider, never alters the original image, and bundles image optimization and accessible alt text on top.
The key differences
| IMG Performer | IP2C | |
|---|---|---|
| Vendor | Enfants Terribles digital GmbH (Hamburg) | Raphael Ehm Mediendesign |
| Processing | image optimization local, on your own server | partly external (visual detection via US vision AI) |
| Account required | no | – |
| AI model (alt text) | your choice – Mistral (France) recommended | – |
| Billing | license, no quota | – |
| Data protection | image optimization without transfer to third parties; for the optional alt-text AI the image goes to your AI provider (DPA with that provider) | – |
| EU AI Act Art. 50 | yes, already in the free version | yes |
| WebP / AVIF | yes (AVIF in Pro) | no |
| AI alt text | yes (Pro, your own API key, no credit cap) | no |
| AI image detection | automatic via C2PA/XMP provenance data, otherwise manual classification | C2PA/IPTC + optical AI detection (GPT-4o, US) |
| Marker in all image sizes | yes – machine-readable marker in every size incl. WebP/AVIF | burns label into a new image file, original is altered |
| Visible label | yes, customizable design (Pro) | yes (3 designs) |
| Platforms | WordPress | – |
| Cost | Free · Pro €69/year or €9/month | – |
Last verified: September 2026. The current vendor pages are authoritative.
What IP2C does well
IP2C (by Raphael Ehm Mediendesign) combines three detection paths: C2PA manifests, IPTC/XMP metadata, and optical image recognition via a vision AI model. That’s more than pure metadata detection — images without (or with stripped) metadata can still be identified as AI-generated. On top of that come three ready-made badge designs, customizable text, and an automatic scan of the existing media library.
Where it falls short
The optical detection runs on a US vendor’s vision AI model (GPT-4o/GPT-4o-mini) — any image that isn’t already unambiguous via metadata goes to the US for analysis. For GDPR-sensitive projects, that’s a point to clarify before deployment.
Technically, IP2C writes the label directly into a new image file (“burned in”), not as a swappable metadata entry. That has an upside (survives screenshots, downloads, social media uploads) and a downside: the original image is altered, and a later correction means a new image.
And: IP2C is a pure labeling tool. Image optimization (WebP/AVIF) or AI alt text for accessibility aren’t part of the feature set — that requires a second plugin.
How IMG Performer solves it
IMG Performer also detects AI images via C2PA manifests, IPTC/XMP metadata, and generator signatures — entirely locally, without any image leaving the server. The labeling lands as a machine-readable marker in every generated image size, including WebP and AVIF — the original stays unchanged, and a correction is always one click in the media library, not a new image.
Anyone who wants a custom label design can upload their own image in the Pro version (not just pick text/color) — and the same Pro version also scans the existing media library retroactively.
The difference: image optimization and alt text aren’t a separate plugin, but the same building block. WebP/AVIF conversion is already included in the Free version, AI alt text for accessibility compliance comes with the Pro version — with your own API key, no credit cap.
Disclaimer
This comparison is based on publicly available information about IP2C (as of September 2026). Features and prices can change — when in doubt, check the vendor’s own page.
Frequently asked questions
Do I need both plugins at the same time?
No — both label AI images, so running them in parallel would only overlap. The decision comes down to whether IP2C's additional optical detection matters more to you than local processing without a US transfer.
Is IP2C's optical detection strictly necessary?
No — for most AI generators (DALL-E, Firefly, Imagen, current smartphone cameras) metadata detection is enough, since the images already carry a C2PA manifest. Optical detection mainly helps with images whose metadata was stripped.
What if I already use IP2C and want to switch?
The images themselves are already altered by IP2C (label burned in) — reverting to the original is only possible with your own backup. New uploads can be labeled directly with IMG Performer from the switch onward.