Masked Diver Battles Eyeless Leviathan in Flooded Ruins
A cinematic underwater action sequence follows a masked diver in tactical gear as she is dragged through flooded stone corridors by a colossal eyeless leviathan. The camera plunges and spirals with her as she fires a harpoon into the creature's jaw, ending with a burst through the surface of an underground ocean lit by bioluminescent organisms.
Dheepan Ratnam @Dheepanratnam
Prompt (original, unmodified)
A flooded subterranean megastructure, ancient stone corridors half-submerged in black water, bioluminescent algae pulsing along the walls like a heartbeat. A masked diver in scarred tactical gear, lit by a flickering wrist-mounted flare, is pulled downward through collapsing flooded chambers by the current of a colossal eyeless leviathan moving beneath the floor grates. The camera plunges after the diver as water rushes through shattering stone archways, air bubbles and debris spiraling past the lens. The diver grabs a broken pillar to stop herself, the current ripping at her body, then releases and freefalls through a vertical shaft, twisting to dodge the leviathan's translucent tendrils whipping upward through the darkness. Quick shaky frames catch her igniting a harpoon gun mid-fall, the flare illuminating the creature's massive jaw opening below. In the final movement, the camera spirals downward around her as she fires the harpoon into the creature's mouth, the bioluminescence erupts blinding white, and the camera bursts through the surface of a vast underground ocean, the diver silhouetted against the glow of a thousand pulsing organisms on a cathedral-sized cavern ceiling. 15 seconds..
Reproduced verbatim from the original post. Never edited or translated. Everything outside this block is our own commentary.
How this prompt compares
The techniques this prompt uses are more common in cinematic AI prompts than in the library overall:
| Technique | In cinematic AI | Library average | Lift |
|---|---|---|---|
| orbit | 20% | 14% | 1.48x |
| push-in | 15% | 12% | 1.31x |
| volumetric | 37% | 23% | 1.63x |
| silhouette | 11% | 8% | 1.35x |
Across all 5,926 prompts, the techniques used here appear at: tracking 41%, handheld 24%, volumetric 23%, low-key 17%.
The 15s duration is not arbitrary: 71% of the prompts in this library that state a duration ask for exactly 15 seconds — Seedance 2.0 accepts clips up to 15 seconds, so 15s is the model's ceiling and the most common choice.
