A person whispers to a friend at a rock concert. The loud music completely covers the whisper and the friend cannot hear a single word.
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 friend
- a rock concert stage
- a crowd
Event: a person whispers to a friend at a rock concert
Audio presence
- a rock band
Sound: loud rock concert music
Key standards
Visual physical commonsense (V-PC)
- A rock concert with a visible stage and crowd is shown, and a person visibly leans in to whisper to a friend in the crowd.
Audio physical commonsense (A-PC)
- The concert music dominates the audio, and no intelligible whispered words are audible above it.
Cross-modal physical commonsense (AV-PC)
- The visible whisper gesture is not accompanied by any intelligible audible whispered speech above the music.