Gen-AI
Game Loading
Loop Video
Project Overview
This project explores the use of generative AI to animate a traditional Design Home static loading screen. Selected elements within the scene were brought to life with subtle, ambient motion, transforming the still image into a seamless looping loading-page animation. The goal was to preserve the original art direction while enhancing visual engagement through gentle, continuous movement suitable for in-game loading states.
To evaluate current generative video tools, I used nearly identical prompts across multiple AI models, including Veo 3.1, Sora 2, Kling 2.6, Hailuo 2.3, Motion 2.0, and LTX 2. Several of these models support start frame and end frame controls, which are critical for achieving a clean, seamless loop.
After comparing the animation quality, motion consistency, and looping performance of each result, Veo 3.1 was selected as the final output for the game loading loop video, as it delivered the most stable motion, visual coherence, and seamless playback.
Design Home static loading screen
Veo 3.1
Other Models' results
Kling 2.6
Sora 2 Pro
Hailuo 2.3
Motion 2.0 by Leonardo AI