Denoising
Denoising is the core operation of a diffusion model: at each step it predicts and removes a little noise, gradually turning a random field into a clear image.
Denoising is exactly what it sounds like: removing noise. A diffusion model generates by starting from random noise and denoising it across many steps. Each step the model estimates "what noise is in this image right now" and subtracts a portion of it, so the picture sharpens into focus a little more each pass.
Denoising strength in img2img
In image-to-image, you also control denoising strength - how much noise is added to your source image before the model starts cleaning it up. A low strength keeps the original mostly intact; a high strength noises it so heavily that the model effectively reinvents it. This single dial is what lets you choose between a subtle edit and a dramatic transformation.
Why it matters
Denoising is the literal mechanism of generation, so it connects to everything else: the sampler decides how each denoising step is computed, the number of steps decides how gradually it happens, and in img2img the denoising strength decides how much of your input survives.
Try it in the generator
Put denoising to work right now - free daily generations, commercial license included.
Related terms
- Diffusion modelA diffusion model is the type of AI that powers most modern image generators. It learns to turn random noise into a coherent image by reversing a step-by-step noising process.
- Image-to-imageImage-to-image (img2img) is the AI workflow that transforms an existing picture according to your prompt, keeping some of the original structure instead of generating from scratch.
- StepsSteps (sampling steps) are the number of denoising passes the model runs to turn noise into an image. More steps can mean more detail, but past a point they only add time, not quality.
- SamplerA sampler is the algorithm that decides how noise is removed at each step of generation. Different samplers reach the final image by different paths, trading off speed, detail and consistency.