Not just generation — we help you earn the click. Prompts follow cover-design rules from the public Fake Bilibili experiment by He Tongxue: a simple layout, text that adds information, and a face that is not always helpful. A YouTube blind test hides view counts until you guess.
Simple layout, text that adds information, and careful use of a face
Two real thumbnails. View counts stay hidden until you guess
Turn a still cover into a short clip for YouTube, Shorts, or Bilibili
Drawn from the public cover experiment by He Tongxue, with no view-count claim
Information overload is hard to scan. A simple cover and a simple title are easier to read
A familiar face can be the brand. A new face can distract, unless the expression is truly striking
Repeating the title adds nothing. Text helps only when it adds a fact or an information gap
Cover and title amplify a video. They do not rescue weak content
Click a scenario card — its prompt fills the generator below
Welcome to ChatIMG Image Assistant
Upload an image or describe the image you want to generate
Click or drag and drop an image
Supports multiple images — AI will blend them together
Two real YouTube thumbnails, view counts hidden. Guess which one earned more clicks.
Enter the blind test →