A sweeping aerial shot of neon towers reflecting on a river at night, captured in cinematic style.Im... - 画廊 - Free Ai Video Generator

提示词: A sweeping aerial shot of neon towers reflecting on a river at night, captured in cinematic style.Image-to-video prompt. Use the provided image as visual reference and style reference, NOT as a single character. Video format: Vertical 9:16, duration 7–9 seconds, 1080x1920, cinematic quality. Scene composition (IMPORTANT): There are EXACTLY 5 brainrot forest creatures in the scene. All creatures are visible at the same time. Character description: Multiple brainrot-style forest creatures (4–6 total), each based on the reference image: – humanoid bodies made of roots, trees and moss – oversized realistic human feet – long exaggerated noses – slightly dumb, calm brainrot facial expressions Each creature looks similar but NOT identical. Small variations in height, nose length, moss density. Positions in frame: – 1 main creature in the center foreground – 2 creatures slightly behind on the left and right – 2 smaller creatures in the background, partially blurred by depth of field All are standing on the forest road. Camera: Low-angle wide shot to fit all characters in frame. Static camera, no shake. Strong depth of field: foreground sharp, background slightly blurred. Lighting: Soft cinematic daylight with volumetric god rays. Consistent lighting across all characters. Natural shadows under each set of feet (IMPORTANT). Animation / motion: 0–2 seconds: All creatures stand still. Only subtle breathing motion. 2–5 seconds: Asynchronous brainrot motion: – one creature slowly tilts head – another blinks – another slightly curls toes – moss and leaves gently move on multiple bodies Movements are slow, minimal, NOT synchronized. 5–8 seconds: All creatures slightly lean forward together, as if noticing the viewer. End in a meme-like frozen pose. Style: Ultra-detailed surreal 3D, meme-core brainrot aesthetics. Hyperreal textures mixed with absurd proportions. Clean, cinematic, TikTok brainrot style. Mood: Absurd, hypnotic, calm, funny. Designed for looping and high retention. Animation quality: Ultra-smooth, no jitter, no character merging. Negative prompt: Single character, duplicated clone overlap, character fusion, missing limbs, extra limbs, fast motion, horror, glitches, text, subtitles, watermarks.. 模型: sora-2. 探索由社区创作并公开分享的 AI 视频作品。

提示词

视频

A sweeping aerial shot of neon towers reflecting on a river at night, captured in cinematic style.Image-to-video prompt. Use the provided image as visual reference and style reference, NOT as a single character. Video format: Vertical 9:16, duration 7–9 seconds, 1080x1920, cinematic quality. Scene composition (IMPORTANT): There are EXACTLY 5 brainrot forest creatures in the scene. All creatures are visible at the same time. Character description: Multiple brainrot-style forest creatures (4–6 total), each based on the reference image: – humanoid bodies made of roots, trees and moss – oversized realistic human feet – long exaggerated noses – slightly dumb, calm brainrot facial expressions Each creature looks similar but NOT identical. Small variations in height, nose length, moss density. Positions in frame: – 1 main creature in the center foreground – 2 creatures slightly behind on the left and right – 2 smaller creatures in the background, partially blurred by depth of field All are standing on the forest road. Camera: Low-angle wide shot to fit all characters in frame. Static camera, no shake. Strong depth of field: foreground sharp, background slightly blurred. Lighting: Soft cinematic daylight with volumetric god rays. Consistent lighting across all characters. Natural shadows under each set of feet (IMPORTANT). Animation / motion: 0–2 seconds: All creatures stand still. Only subtle breathing motion. 2–5 seconds: Asynchronous brainrot motion: – one creature slowly tilts head – another blinks – another slightly curls toes – moss and leaves gently move on multiple bodies Movements are slow, minimal, NOT synchronized. 5–8 seconds: All creatures slightly lean forward together, as if noticing the viewer. End in a meme-like frozen pose. Style: Ultra-detailed surreal 3D, meme-core brainrot aesthetics. Hyperreal textures mixed with absurd proportions. Clean, cinematic, TikTok brainrot style. Mood: Absurd, hypnotic, calm, funny. Designed for looping and high retention. Animation quality: Ultra-smooth, no jitter, no character merging. Negative prompt: Single character, duplicated clone overlap, character fusion, missing limbs, extra limbs, fast motion, horror, glitches, text, subtitles, watermarks.