integrated_multimodal_description: [Shot 1] Handheld cellphone footage from a passenger seat on a commercial airplane, filmed through the window during daylight above a thick cloud layer. Mild turbulence, realistic cabin ambience, totally normal amateur footage at first: the wing, clouds, and slight shaking from the aircraft. From 00:00.000 to 00:03.500, nothing unusual happens. At 00:03.500 the cloud deck far below begins to bulge and part in a disturbing way. At 00:05.000 a colossal dark tentacle, thicker than a skyscraper, slowly rises out of the clouds and reaches upward toward the plane. The filmer’s voice cracks, [English, terrified] Oh my god— what is that?!. At 00:07.500 the tentacle suddenly accelerates toward the aircraft. At 00:09.000 it smashes violently through the fuselage and into the cabin, causing a burst of debris, wind, screaming passengers, and structural tearing. But instead of attacking further, the end of the gigantic tentacle is revealed to be holding an enormous paper takeout bag. From 00:10.000 to 00:14.500 the camera shakes wildly amid the damaged cabin as the tentacle awkwardly extends the bag deeper into the aisle like a delivery driver completing a dropoff. A calm, matter-of-fact voice from somewhere in the clouds says, [English, calm] DoorDash for Kevin?. The contrast should be completely deadpan and unexpected, with no comedic music or visual cue beforehand. overall_soundscape: Realistic airplane cabin noise, muffled engine hum, mild turbulence, rising passenger panic, violent impact, metal tearing, rushing wind, screams, rattling debris, frantic breathing from the filmer, followed by the absurdly calm delivery line from outside. non_diegetic_music: none
This clip was generated with MiniMax H3 by MiniMax. The prompt on this page is the exact text used to create it.
The example was generated with MiniMax H3 using: 9:5 · 0:15 · Cinematic · Horror · Mysterious. Reusing these settings gives the closest match.
Copy the prompt as-is first, then change the subject while keeping the camera, motion, and lighting terms. Video models are sensitive to wording, so expect a few runs.
Yes. Prompts work best as starting points — swap scenes, adjust pacing terms, or change the aspect ratio to fit your project.