Glossary

Text-to-Image

Short answer

Generating an image from a written description; the basic mode of every tool here.

Text-to-image is the default mode of AI image tools: you write a prompt and an image comes out.

Its alternatives are img2img, inpainting (regional regeneration) and ControlNet (structural reference). Those all require a starting image.

Text-to-image is the fastest method and the least controlled. When the result has to take a specific form, you have to move to one of the others.