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.
See also
You will meet this in
- Midjourney — The Aesthetic Quality LeaderThe subscription generator that delivers the most consistent artistic and aesthetic results.
- DALL·E 3 — Generating Images by Talking Inside ChatGPTThe most prompt-faithful generator, steerable through a conversation in ChatGPT.
- Flux — The Model That Leads on Photorealism and TextThe strongest open-weight family for photorealistic people and for text rendered inside images.