Calabi Labs · Guide · 2026-06-14
Ai image resizer
```html
Does Calabi Resize AI Images?
No — Calabi does not resize images or change their dimensions. Calabi is a one-pass forensic cleaner: it strips the invisible AI-detection signals and metadata that platforms scan for, injects authentic phone-capture identity, and gives you a proof card showing exactly what changed. If you need to change width × height, any basic image editor does that. If you need to post your AI image without it being flagged as AI-generated, that's what Calabi handles.
What Actually Gets Your AI Image Flagged
When you export an image from Midjourney, DALL-E, Stable Diffusion, or any other model, the file carries a dense layer of invisible metadata that platform scanners read before your content even appears publicly. This is not visible to viewers — but it is machine-readable, and platforms like Instagram, TikTok, YouTube, and Reddit run those scanners automatically on every upload.
The specific signals that trigger detection include C2PA / Content Credentials — a cryptographic manifest stored as JUMBF atoms that explicitly declares the image was AI-generated, including a reference to the training data. Your raw AI export also carries an XMP tag: DigitalSourceType = trainedAlgorithmicMedia, which is Adobe's standardized flag for AI-sourced content. The encoder fingerprints are another signal: Lavc (libavcodec) and x264 SEI metadata in video exports, or similar encoder markers in image files, tell scanners this file was rendered by a compute pipeline, not captured by a sensor.
Platforms also flag files that lack the metadata profile of a real phone capture — no Make/Model, no GPS coordinates, no capture timestamp, no genuine camera serial. A missing or wrong device identity is itself a red flag in 2026-era detection systems.
Why Resizing Doesn't Help
Changing an image's pixel dimensions — making it 1080×1350 for an Instagram post, 1200×675 for YouTube — does nothing to the invisible metadata layer. The C2PA manifest survives crop and resize. The XMP AI flags survive recompression. The encoder fingerprint is embedded at the byte level and remains intact through dimension changes. Platforms scanning your upload are reading that metadata, not measuring your image's width.
Screenshotting and re-uploading disrupts some visible watermarks but leaves the metadata largely intact in most export paths. Re-encoding through a third-party tool sometimes strips some tags but also tends to degrade quality and still leaves C2PA and encoder fingerprints untouched unless specifically targeted for removal.
How Calabi Actually Cleans Your AI Image
Calabi runs a three-stage pipeline on every upload, fully automatically — no manual settings or technical knowledge required.
- Strip: Calabi removes C2PA / Content Credentials JUMBF atoms, the DigitalSourceType: trainedAlgorithmicMedia XMP tag, generator and tool tags, and encoder fingerprints like Lavc and x264 SEI. A raw AI export carries roughly 144 metadata tags; Calabi reduces that to about 94 neutral structural tags.
- Inject: Calabi writes authentic phone-capture identity into the file — a real device profile (iPhone 15 Pro, Pixel 8 Pro, Galaxy S24 Ultra), real GPS coordinates, a plausible capture timestamp, and a genuine phone encoder name. The file now looks, at the forensic level, like it was shot on a real device.
- Verify: Before download, you receive a forensic proof card — generated by ExifTool, the same tool newsrooms and courts use — showing exactly what was stripped and what was injected. You can read it yourself and see the before-and-after state.
FAQ
Does Calabi remove visible watermarks like Sora's sparkle icon or a corner logo?
Calabi does not edit pixels and cannot erase a visible logo or watermark. What survives cropping and re-upload is the invisible detection layer — C2PA manifests, XMP flags, and encoder fingerprints — which is exactly what Calabi removes. If the visible watermark is a concern, cropping it out in any image editor removes it; Calabi then cleans the metadata layer that otherwise betrays the file's AI origin.
Can I use Calabi to resize images to specific social media dimensions?
Calabi preserves the dimensions of your uploaded file. It does not resize or crop to target dimensions. Use a dedicated resizer or your platform's built-in editor for dimension changes, then run the cleaned file through Calabi to strip the AI-detection metadata before uploading.
Will this guarantee my post won't be flagged?
No tool can guarantee a platform won't flag you — detection methods evolve. Calabi removes the documented metadata and encoder signals that current automated scanners rely on, and it does so comprehensively: 18 JUMBF atoms reduced to 0, 16 C2PA references to 0, the trainedAlgorithmicMedia flag removed. Results vary by platform and source model.
Try Calabi free at calabilabs.com — 10 cleans, no card.
```