Seed
A 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.
Every text-to-image generation starts from a field of random noise. The seed is the number that determines exactly what that "random" starting noise looks like. Because the generation process is otherwise deterministic, the seed is what makes one run different from the next.
Why it matters
- Reproducibility: save the seed and you can recreate an image exactly, even days later.
- Controlled iteration: lock the seed, then change one word or one setting to see its isolated effect.
- Variations: keep the prompt and change only the seed to explore many different takes on the same idea.
How to use it
A "random" seed (often shown as -1) tells the generator to pick a new number each time, giving fresh results on every run. Once you get an image you like, fix its seed and tweak the prompt, CFG scale or steps to refine that specific composition instead of rolling the dice again.
Try it in the generator
Put seed to work right now - free daily generations, commercial license included.
Frequently asked questions
Will the same seed give the same image on a different tool?
Not necessarily. Seeds are reproducible within the same model, sampler and settings. A different generator, model version or sampler can interpret the same seed differently.
Related terms
- 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.
- 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.
- 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.
- CheckpointA checkpoint is a saved AI model file containing the full set of trained weights. It is the complete "brain" that generates images - swapping checkpoints changes the entire look and capability.