A car engine idles. The driver presses the gas pedal gradually, and the engine roar grows steadily 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 car
- a gas pedal
- a driver
Event: a driver presses the gas pedal of a car
Audio presence
- a car engine
Sound: a car engine running
Key standards
Visual physical commonsense (V-PC)
- The car visibly idles first, then the driver visibly presses the gas pedal with gradually increasing depth in a smooth continuous press.
Audio physical commonsense (A-PC)
- The engine sound grows progressively louder as the pedal is pressed further, rising smoothly from the idle level.
- The sound is a real continuous car engine running and revving.
Cross-modal physical commonsense (AV-PC)
- The rising engine loudness aligns in time with the visible depression of the gas pedal.
- The louder engine sound corresponds to a more deeply pressed gas pedal.
- The idle-level engine sound corresponds to a released or barely-pressed gas pedal.