Steps
Steps (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.
A diffusion image is created by gradually removing noise over a series of passes. Each pass is a step. The step count tells the sampler how many increments to use as it travels from pure noise to a finished image.
How it affects the image
- Too few steps (under ~10 for many samplers): the image can look unfinished, smudgy or noisy.
- Enough steps (~20-35 for most modern samplers): clean, detailed results.
- Too many steps (60+): diminishing returns - the image barely changes while generation time keeps climbing.
Why it matters
Steps are a direct trade between speed and quality. The right number depends heavily on the sampler: some efficient samplers produce a great image in 4-8 steps, while older ones need 30+. When you raise steps, you usually want to keep the seed fixed so you are comparing the same image at different step counts.
Try it in the generator
Put steps to work right now - free daily generations, commercial license included.
Frequently asked questions
How many steps should I use?
For most modern samplers, 20-30 steps is a solid default. Fast "turbo" or distilled models can look great in just 4-8 steps. More steps rarely improve a good image - they mostly cost time.
Related terms
- 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.
- CFG scaleCFG scale (classifier-free guidance scale) controls how strongly the image follows your prompt. Low values are loose and creative; high values stick closely to the prompt but can look over-processed.
- SeedA seed is the number used to initialize the random noise an image is built from. The same seed plus the same prompt and settings produces the same image every time, which makes results reproducible.
- DenoisingDenoising 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.