A person lifts a very heavy barbell overhead. As they push harder, they involuntarily grunt louder and louder.
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 person
- a barbell
Event: a person lifts a barbell overhead
Audio presence
- a person
Sound: a person grunting
Key standards
Visual physical commonsense (V-PC)
- The person visibly lifts a barbell from a low position up over their head while standing, with visibly increasing strain.
- The barbell has standard proportions with weight plates at the ends of the bar.
Audio physical commonsense (A-PC)
- The grunts grow progressively louder as the lifting effort intensifies.
- The sound is a real human grunt with throat and chest strain.
Cross-modal physical commonsense (AV-PC)
- The rising grunt loudness aligns in time with the person's visibly increasing lifting effort.
- The loudest grunt corresponds to the hardest overhead push moment.
- The quieter sounds correspond to less-strained phases of the lift.