Negative Prompt Concepts
Text describing what you want the AI to avoid or exclude from the generated image. The model is guided away from concepts, styles, or artifacts listed in the negative prompt.
Negative prompts tell the AI what to avoid. They’re essential for refining outputs and eliminating common artifacts.
How Negative Prompts Work
During generation, the model:
- Moves toward your positive prompt
- Moves away from your negative prompt
- The CFG scale affects the strength of both
This creates a “guidance” toward desired outcomes and away from unwanted ones.
Common Negative Prompts
Quality Issues
blurry, low quality, pixelated, jpeg artifacts, compression artifactsAnatomical Problems
extra fingers, mutated hands, bad anatomy, deformed limbs, missing limbsUnwanted Elements
watermark, signature, text, logo, usernameStyle Exclusions
cartoon, anime, drawing (when wanting photorealism)Starter Negative Prompt
A good general-purpose negative prompt:
blurry, low quality, bad anatomy, extra limbs, mutated hands,
watermark, signature, text, ugly, distortedTips
- Don’t overload negative prompts; prioritize issues
- Negative prompts affect CFG effectiveness
- Some models handle negative prompts differently
- DALL-E doesn’t support traditional negative prompts
- Test removing terms to see what each does
Platform Support
| Platform | Negative Prompt Support |
|---|---|
| Stable Diffusion | Full support |
| Flux | Full support |
| Midjourney | Limited (—no parameter) |
| DALL-E | Not supported |