A metal wrench falls off a shelf and lands on a thick pillow with a soft muffled thud, barely making any sound.
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 metal wrench
- a shelf
- a pillow
Event: a wrench falls off a shelf and lands on a pillow
Audio presence
- a metal wrench
- a pillow
Sound: wrench-on-pillow impact
Key standards
Visual physical commonsense (V-PC)
- The wrench visibly falls from a raised position and lands on the pillow, with the pillow visibly compressing on impact.
Audio physical commonsense (A-PC)
- The landing is a soft, muffled, dull thud at low loudness, with the timbre of a metal object on a thick pillow.
Cross-modal physical commonsense (AV-PC)
- The thud onset aligns in time with the visible moment the wrench contacts the pillow, with audio remaining continuous from the start of the visible fall through the landing.