Negative prompt
A negative prompt is a list of words describing what you do NOT want in the image. The model steers away from those concepts while it generates, which removes common flaws and unwanted elements.
Where your main (positive) prompt describes what you want, a negative prompt describes what you want kept out. The generator pushes the image away from the concepts in your negative prompt at every step, which is a fast way to fix recurring problems without rewriting the whole prompt.
How it works
Technically, the model is guided in two directions at once. The CFG scale blends a prediction conditioned on your positive prompt with one conditioned on the negative prompt, and the final step is nudged toward the positive and away from the negative. Higher CFG makes both the positive and negative pull stronger.
What to put in it
- Quality flaws: "blurry, low quality, jpeg artifacts, lowres, watermark, text"
- Anatomy issues for people: "extra fingers, deformed hands, mutated, bad anatomy"
- Unwanted content or style: "cartoon" (if you want photoreal), "people" (if you want an empty scene)
Keep it focused. A bloated negative prompt can strip out detail you actually wanted, so add terms in response to problems you actually see rather than pasting a giant generic list.
A clean, reusable negative prompt for photorealistic portraits that removes the usual artifacts.
Example prompt
blurry, lowres, jpeg artifacts, watermark, text, deformed hands, extra fingers, bad anatomy, oversaturatedTry it in the generator
Put negative prompt to work right now - free daily generations, commercial license included.
Frequently asked questions
Do I always need a negative prompt?
No. Modern models produce good results without one. Reach for a negative prompt when you keep seeing a specific problem - extra fingers, a watermark, the wrong style - and want to suppress it.
Related terms
- Prompt weightingPrompt weighting lets you boost or reduce the influence of specific words in your prompt, so the model pays more attention to the parts that matter most to you.
- 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.
- Text-to-imageText-to-image is the AI workflow where you type a written prompt and the model generates a brand-new image from it - no source image required.