AI Glossary / Text-to-image
🎨 Image, video and audio
Text-to-image
Generating images from text descriptions (Midjourney, DALL·E, Nano Banana).
Text-to-image is generating images from text descriptions: you type "astronaut cat in watercolor style," the model creates such an image. In a few years the technology went from blurry curiosities to graphics indistinguishable from photographs and professional illustrations. Tools differ in strengths: some excel at aesthetics, others at photorealism, still others at rendering text in images or editing existing photos. The common denominator is prompt control - and increasingly also conversation, where successive commands refine the same image.
Related concepts