What are Negative Prompts?

Negative prompts tell the AI "what NOT to draw" — commonly used to filter out picture flaws and unwanted elements.

Auto-applied default negative prompts

The system ships a built-in default negative prompt per model, covering the most common picture issues (artifacts, low quality, watermarks, distortions). These don't need to be typed manually — they're applied to every generation automatically.

Basic model defaults

blurry, lowres, upscaled, artistic error, film grain, scan artifacts,
worst quality, bad quality, jpeg artifacts, very displeasing,
chromatic aberration, halftone, multiple views, logo,
too many watermarks, negative space, blank page

Unrestricted model defaults

nsfw, lowres, artistic error, film grain, scan artifacts,
worst quality, bad quality, jpeg artifacts, very displeasing,
chromatic aberration, dithering, halftone, screentone,
multiple views, logo, too many watermarks,
negative space, blank page

TIP

You don't need to repeat any of the above — the system already adds them for you. If the defaults already produce good results, leave the negative prompt field empty.

When you want to exclude something extra

If the defaults don't cover your specific case, add to the negative prompt field. Common cases:

Want to avoid	Add
Weird fingers	bad hands, extra fingers, missing fingers
Text watermarks	text, watermark, signature
A specific element	Write the element directly, e.g. glasses (no glasses), hat (no hat)
Multiple people	multiple people, crowd
Blurred background	blurry background, bokeh

Tips

• Don't repeat: don't re-add what's already in the defaults
• Add as needed: too many negative terms confuse the AI — don't add anything if you're not seeing issues
• Reverse-engineer from problems: look at what's wrong with the actual output, then add a targeted negative term
• Final effective content = default negative prompts + your additions