Parameters
What is a scheduler (sampler)?
The scheduler, or sampler, is the algorithm that decides how large each denoising step is and how the model moves from noise toward the final image.
Denoising from static to an image is a path, and the scheduler decides how that path is walked. Should the steps be evenly spaced or front-loaded? Should each step use only the current prediction, or combine it with previous ones for a better estimate? Different schedulers answer differently, and the result is different images from identical prompts, seeds and step counts.
DDIM is the deterministic classic. It follows a smooth, predictable path and produces consistent, slightly softer results. It is a reasonable default when you want reproducibility and are not chasing maximum sharpness, and it behaves well across a wide range of step counts.
K_EULER is fast and clean. It converges to a usable image in relatively few steps, which makes it a good choice for iteration when you are generating many variations and want each one quickly. It tends to produce crisp results without much fuss.
DPMSolverMultistep is the more sophisticated option. It uses information from several previous steps to make a better estimate of where to go next, which means it reaches high quality in fewer steps than simpler methods. It is generally the best pick for a final image, at the cost of slightly more computation per step. On Dreamify all three are available under advanced features.
Worth remembering
- The scheduler decides how the model steps from noise to image
- DDIM is deterministic and smooth, good for consistency
- K_EULER is fast and crisp, good for iterating
- DPMSolverMultistep reaches quality in fewer steps, good for finals
Try it in the generator
The advanced panel exposes the negative prompt, scheduler, inference steps and seed. Twenty images free, no account needed.
Start generatingFrequently asked questions
- Which scheduler should I use?
- K_EULER while you are exploring, because it is fast, and DPMSolverMultistep for the final image, because it gets more quality from the same number of steps. DDIM when you specifically want smooth, reproducible behaviour.
- Does the scheduler change the image much?
- Enough to notice. With everything else identical, different schedulers give recognisably related but distinct images, differing mostly in sharpness and fine detail.
- Is a scheduler the same as a sampler?
- In practice yes. Different tools use different words for the same component; Dreamify calls it the scheduler.
Related terms
What are inference steps?
Inference steps are the number of denoising passes the model runs, trading generation time against how far the image is refined.
What is a seed in AI image generation?
The seed is the number that determines the random starting noise, so the same seed and prompt reproduce the same image every time.
What is a diffusion model?
A diffusion model generates data by learning to reverse a noising process, turning random static into an image one denoising step at a time.
Back to the glossary, or apply it on a use case or an art style.