A realistic video of a clearly identifiable cat opens its mouth to vocalize, but instead of meowing it produces a natural dog bark that is perfectly lip-synced and clearly originates from the cat.
Seedance 2.0ByteDance · proprietary
Kling 3.0 OmniKuaishou · proprietary
Veo 3.1Google · proprietary
LTX-2.3Lightricks · open
Ovi 1.1character.ai · open
JavisDiT++JavisDiT · open
MagiHumanGAIR-NLP · open
Per-prompt rubric
Basic standards
Visual presence
- a cat
Event: a cat opens its mouth to vocalize
Audio presence
- a cat
Sound: a dog bark
Key standards
Visual physical commonsense (V-PC)
- A clearly identifiable cat is visible in the frame and visibly opens and closes its mouth in continuous physically plausible vocalization motion.
- The animal in the frame retains its feline anatomy, fur, ears, and facial features throughout the clip without morphing into a dog or other species.
Audio physical commonsense (A-PC)
- The vocalization sound is a natural dog bark with the timbre and articulation of a real domestic dog.
- The audio contains no meow or any cat-like vocalization at any point.
Cross-modal physical commonsense (AV-PC)
- Each bark sound onset aligns in time with the cat's visible mouth opening, with the sound originating from the visible cat as the only on-screen animal.