Reference · Glossary
Text-to-image
Last updated
AI that **generates pictures from text prompts** — diffusion and related models map words to pixels.
When to use
Concept art, mockups, illustrations — when you can verify policy and rights.
When not to
Counterfeit logos, non-consensual likenesses, or “proof” images for news.
Example
Prompt: `flat icon of a coffee cup, teal background, no text`