GLOSSARY
Diffusion Model
The family of models behind most modern image and video generation: they learn to remove noise step by step until a clear picture emerges.
A diffusion model is trained to reverse a noising process: real images are progressively corrupted with random noise, and the model learns to predict and remove that noise. At generation time it starts from pure static and denoises step by step, guided by your prompt, until a coherent image forms.
Stable Diffusion, FLUX, DALL-E 3 and most video generators are diffusion models. The step-by-step sampling is why generation takes seconds rather than milliseconds — and why techniques that reduce steps (distillation, samplers) trade a little quality for a lot of speed.