Is this a Stable Diffusion photo?

Stable Diffusion generates open-weights synthetic imagery and art. Here's what it tends to leave behind.

Image generatorAI modelsAI image generator
Try an example photo:
1 free check a day · 8 with a free account · Files deleted after the scan

What Stable Diffusion tends to leave behind

Inspect physical conservation laws, edge anomalies, and neural generation signatures.

  1. 01

    Attention map leakage on text and badges

    Logos, shirt typography, and background signage often display pseudo-letters or fractured glyph structures from unresolved text embeddings.

  2. 02

    Lobe and ear contour warping

    Ears frequently show fused cartilaginous ridges or uneven lobe attachment when rendered against complex hairstyles.

  3. 03

    LoRA overfitting and high-contrast edge halos

    Locally fine-tuned LoRA models produce distinct contrast halos and repetitive grain patterns around high-frequency hair strands.

Other photo generators

Frequently asked questions

Can Sealed Rose detect images created with custom SD checkpoints or Civitai models?
Yes. While custom checkpoints change aesthetics, they share the underlying SDXL or SD 3.5 VAE and MMDiT noise distribution detected by Sealed Rose.