Skip to content
  1. Frame 0
  2. Frame 10
  3. Frame 20
  4. Frame 30
  5. Frame 40
  6. Frame 51
  7. Frame 61
  8. Frame 71
  9. Frame 81
  10. Frame 91
  11. Frame 102
  12. Frame 112
  13. Frame 122
  14. Frame 132
  15. Frame 143
Frame 015 of 144 sampledFrame 143

latte-art-in-a-cafe.mp4

Example report. Made with Veo 3, so the right answer is Veo 3.

This clip most closely matches Veo 3.

87.1%

Confidence: the probability the model gives Veo 3.

Real
9.5%
Sora 2
3.4%
Veo 3
87.1%
Compared against
Real, Sora 2 and Veo 3
Frames analyzed
15 of 144, evenly spaced
Technical details
The frame as uploadedThe same frame as the model receives it: 360 px high, blurred, JPEG quality 55, 224 by 224
Frame 71, as uploadedAs the model sees it, squeezed to 224 × 224
Returned class
Veo3
p(class)
0.8714
Features
EfficientNet-B0, ImageNet weights, 1280 values per frame
Classifier
SVM with probability outputs over Real, Sora 2, Veo 3
Preprocessing
360 px height, 3×3 Gaussian blur, JPEG quality 55, 224 × 224
Inconclusive
below 0.60
What the file is
Made with Veo 3

Other examples