Describe the scene. Click the rest.
Pick camera, lens, light and style visually, and get a prompt that works with Midjourney, Stable Diffusion, Nano Banana Pro, GPT Image and Flux.
Kadar writes text prompts. It does not generate images; you paste the prompt into the image tool you already use.
Describe only what's in the picture. Any language works, and “Refine with AI” will translate it.
Style
The biggest influence on the look.
Facial expression
For people and characters.
Camera body
What the photo was taken with.
Shot size
How much of the subject we see.
Camera angle
Where the camera looks from.
Lens
Field of view and perspective.
Depth of field
How blurry the background is.
Motion and shutter
How movement is captured.
Lighting
Shape, direction and color of light.
Detail and texture
Pick any. Small details sell realism.
Palette
Dominant colors and tone.
Composition
Where the subject sits in the frame.
Aspect ratio
The shape of the image.
Things the model should leave out.