How to Use Negative Prompts In AI Art – With Examples


If you click on a link and make a purchase, I may receive a small commission. As an Amazon affiliate partner, I may earn from qualifying purchases.
Read our disclosure.

Key Takeaways

  • Negative prompts remove objects, items, or things from the image generation.
  • Negative prompts can be used to modify the image for a more accurate result.
  • The first words have the biggest effects, and the later words are for fine-tuning the text prompt.
  • You can emphasize negative text prompts with brackets.
  • Midjourney has different parameters for controlling negative text prompts.
  • Stable Diffusion v1.5 does not utilize negative prompts the same way as SD v2.x does.

Modifying Image With Negative Prompts

Negative prompts are mostly used to modify the image to your liking. While you can type a lot of words and things into the text prompt field, AI might still create an image that is not to your liking.

By typing keywords into the negative text prompt field, you can modify the next creation to be more aligned with what you are looking for. Think of the negative text prompt as an extra (more precise) guide for the AI to work with.

AI art generated with Getimg by using negative prompts.

Text prompt used (without negative prompt):

(masterpiece best quality),1girlsolo, flower, long hair, outdoors, letterboxed, school uniform, day, sky, looking up, short sleeves, parted lips, shirt, cloud, black hair, sunlight, white shirt, serafuku, upper body, from side, pink flower, blurry, blue sky, depth of field

The above text prompt created two images that had a girl with brown hair but not black hair. As black hair was included in the text prompt, the AI created images that were different than what was expected.

Adding a negative text prompt: brown hair

After adding the negative text prompt brown hair, the AI algorithm generated an image of a girl with black hair.

Sharpening and blurring

Positive prompt: blurry, on the left, and negative prompt: blurry on the right.

While we have defined blurry in the text prompt, we could also define sharp and crisp and write blurry to the negative prompt field. This would result in a sharp image as there are three words pointing to it: positive: sharp, crisp, and negative: blurry.

You can also write blurry to the normal text prompt and sharp, crisp as a negative prompt to signal the AI algorithm that you most definitely want to see a blurred image.


When comparing the two images, a more cartoony vibe is visible on the left side image, whereas the right side image has a more realistic style. Notice the shirt and eyes.

When you want to make the image more photorealistic than cartoony, you can use a negative prompt like: (cartoon: 1.5) to signal the AI that you are looking for a more photorealistic result rather than a cartoon-like result.

Removing Things With Negative Prompts

You can use negative prompts to make the AI not include items, objects, things, or certain elements in the image.

By further defining that I didn’t want to see a ribbon, the AI removed the ribbon and also changed the skirt color to blue, as per requested. AI doesn’t always honor the text prompts and might still include things that you have mentioned not to include.

Negative text prompt used: brown hair, black ribbon, pink dress, red sclera

AI art generated with Getimg by using negative prompts to erase things from the image.

Using Negative Prompts In Midjourney

Midjourney has different parameters to negative prompts as many other AI art generators. In Midjourney, you use “–no” to signal the AI that you don’t want to see certain things. You can also use weighting in Midjourney. Type :: and then the weight at which you want to emphasize a certain thing.

Example: –no 3d render::1.3 -this means that the image should not have a 3D rendering in it with a 30% emphasis on it.

You can add multiple parameters to a single prompt. You can read more about how to use different parameters in Midjourney through their documentation.

Emphasizing The Negative Prompt

If you are using Stable Diffusion (SD) (might also apply to some AI art generators besides SD), you can use brackets () to define the weight that specific prompt has. Brackets work for the positive/normal prompts the same way they work for negative prompts.

Example: (brown hair: 1.4) – this means the AI puts 40% more emphasis on not including brown hair in the image.

While you can use keyword emphasis in Stable Diffusion, especially with (AUTOMATIC1111 WebUI), it might not be supported by all services.

It’s also worth noting that the first words in the text prompt (positive and negative) have the most effect on the end result.

Stable Diffusion And Negative Prompts

While many latest AI art generators have included negative prompts to their platforms as a way to further define what users want to see and mainly don’t want to see, Stable Diffusion v1.5 and v2.x versions act very differently.

In Stable Diffusion 1.5, the base model uses OpenAI’s CLIPText, whereas Stable Diffusion 2.x uses OpenCLIP. Architecture-wise, the models are pretty much the same, but the way the text encoder works is vastly different.

In SD 2.x, the negative prompts have better execution functionality as the model was trained differently than in SD 1.5.

How to Create A Negative Prompt

While you can always type individual words to make an image better, you can also create a universal negative prompt to be used with every single image you create.

To create a universal negative prompt, you can use words like disfigured, ugly hands, ugly face, distorted face, ugly eyes, extra hands, extra limbs, ugly, poorly drawn hands, deformed, body out of frame, bad anatomy, cut off, bad art

Now, whenever you create an image, you can use your pre-typed negative text prompt to see if it will improve the results.

Creating a negative prompt is all about fine-tuning the image and emphasizing what you don’t want to see in an image. It’s up to the AI to honor your request as sometimes you might get results that you have specified as unwanted.

Why Use Negative Prompts

Negative prompts help you fine-tune the image to include only things you want to see in the image. Negative text prompts also help modify the image to have a certain style (crisp, clear, purple, etc.) and also to help correct usual mistakes such as disfigured hands, extra fingers, extra hands, ugly eyes, etc.

AI art generators are still in their infancy, and they make mistakes all the time. While you can get superb results by using just the main text prompt, sometimes the AI adds things to the image that you haven’t specified in the text prompt.

This is the reason why negative text prompts are so effective. These text prompts help you further define what you don’t want to see in the image.



Digital Artist

I’m a digital artist who is passionate about anime and manga art. My true artist journey pretty much started with CTRL+Z. When I experienced that and the limitless color choices and the number of tools I could use with art software, I was sold. Drawing digital anime art is the thing that makes me happy among eating cheeseburgers in between veggie meals.

More Posts

Thank You!

Thank you for visiting the page! If you want to build your next creative product business, I suggest you check out Kittl!