Prompt-only generation lets customers create art from text without uploading a photo. The customer writes a description, chooses a style, sets available parameters, and starts the generation. This mode is useful for imaginary scenes, abstract posters, mood-based artwork, nursery art, interior concepts, and ideas where no personal photo exists. A strong prompt usually includes subject, setting, mood, medium, colors, and composition. The model may still interpret the prompt creatively, so iteration is normal. Prompt-only generation can require tokens, especially for advanced styles. If a customer wants a person, pet, or object to resemble a specific reference, they should use reference images when available and mention them clearly in the prompt.