Hyper-realistic Action Video Generation from Image Reference
This complex prompt is designed for video generation using Google Veo 3.1 and Nano Banana Pro, instructing the AI to create a hyper-realistic action sequence based on an uploaded image reference. The video features a monster charging and striking a Monster Energy can with an axe, with strict constraints on camera movement and scene setup.
Prompt
Create a hyper-realistic action video using the uploaded image as the exact visual reference. SCENE SETUP: A Monster Energy can stands perfectly upright at the center of the frame. The environment stays dark, green, and fog-filled as in the reference image. A large humanoid monster is visible in the background, partially obscured by green fog. CAMERA: Camera is completely locked. Same angle throughout the entire video. No zoom. No pan. No rotation. No camera shake. ACTION SEQUENCE: โ The video starts with stillness. The can does not move. The monster is visible but stationary in the background. โ The monster suddenly charges forward at full speed toward the can. Running motion is aggressive, fast, and direct. No slow motion at any point. No pauses. โ The monster holds a heavy axe with both hands while running. The axe is raised as it approaches the can. โ The monster strikes the can with full force using the axe. The impact is violent and instantaneous.
Open in Hitboard Original post