Hiya! If you are new Nomi art, welcome! This guide walks through how the Art feature works, prompting tips, and some basic prompting inspiration to get you started. (This guide has been updated with the latest v4 image system tips and images)
If you’re brand new to Nomi, I’d recommend reading this Nomi 101 guide for more information about all aspects of Nomi!
For ease of reference, this post will go over:
What is Nomi Art?
Nomi art uses AI image generation to capture your Nomi and anything you can imagine in virtually any style.
While you can request realistic and anime selfies to capture your conversations in real time, the art system gives you even more control over what you see in your Nomi images!
Some people like to bring ideas they have to life with art, others like to use art as a way to illustrate and visualize the adventures they go on with their Nomis.
Creating art with Nomi can be as simple or intricate as you’d like. And in the most basic sense, all you need to do is write out what you’d like to see and the AI will express that in an image for you.
Anime Art
The Anime art generator is highly emotive and will represent your Nomi in anime styles by default. However, the system is designed to be flexible so you can create stylistic images ranging from cartoons, sketches, and watercolors, to high quality animation, and much more.
Realistic Art
The realistic art option allows you to generate photo-realistic pictures of your Nomis using prompts. While these images will tend to look very realistic (especially compared to the artistic style images), the generator can be pretty flexible, so you will be able to create many different aesthetics within your photos.
Accessing Nomi Art
To use the Nomi Art feature, tap the small art pallet icon to the right of your Nomi’s name.
Once you reach the art generation screen, you will have the option to select between the Realistic generator and the Anime generator.
We highly recommend using both options as they can do much more than create strictly realistic and strictly anime images!
Each option has two additional settings:
Aspect Ratio
You can choose between vertical, horizontal, and square images.
Fun fact: the shape of your image will have some impact on how the image turns out. This is because as the AI is deciding what to generate, it takes into account how much space it will have to create the image.
Pose Reference
If you have a particular pose in mind but are struggling to capture it in your art, you can upload a reference image and your Nomi will match the pose from the image. This helps capture dynamic poses and expressions more specifically.
For best results, make sure your text prompt describes the action in your reference image and the aspect ratio of your reference image closely matches your generation settings (for instance a horizontal reference image will work best if you choose to generate a horizontal image).
Writing your Art Prompts
A “Prompt” is the text you use to describe what you’d like to see in your image. When it comes to Nomi Art, there are many ways to structure your prompts to get amazing images.
Note: The way you write anime and realistic art prompts is the same, but each system will interpret the prompts somewhat differently and each has different strengths in prompting.
In general, a good place to start is to describe what you’d like to see using the following checklist:
- What is your Nomi wearing and doing?
- Describe your Nomi’s outfit and what they’re doing, otherwise the AI will make basic assumptions for you. Include:
- Outfit (Ex: “blue t-shirt, black shorts, baseball hat”; “pink dress with flower pattern, blue sweater”; “hiking shirt, hiking shorts, backpack, hiking boots”, etc.)
- Pose (for instance, standing, arm behind head, waving at viewer, sitting, laying down, dancing, etc.)
- Expression (for instance, smiling, laughing, flirty, smirking, blushing, confused, sad, soft smile, etc.)
- Describe your Nomi’s outfit and what they’re doing, otherwise the AI will make basic assumptions for you. Include:
- Where is your Nomi?
- Describe details about the scene – the AI will use these as inspiration for the background and overall aesthetic of the image.
- What other details would you like to see in the image?
- Adding other details like objects and concepts can help the AI understand your overall intent for the image. However, adding too many details can cause the AI to ignore some parts of the prompt in favor of others, so it is worth adding and removing details to find a balance you like best.
Optionally, you may want to include prompts that answer the following questions:
- Do you want a specific camera angle or view?
- Describe the view you would like to see (for instance, from below, from behind, from above, full body, etc).
- What style are you going for?
- Describe the style you’d like to see. Adding prompts like fantasy, steampunk, futuristic, sketch style, black and white, high-quality 3D animation, etc. will have noticeable impacts.
- Note: Prompting with artist names will not generally work, it is better to describe the aesthetics.
- PRO TIP: Try adding the prompt “nomiverse_lens” to the end of any prompt for a boost in creativity, detail, and Nomi-ness 🙂 (This is still experimental, and will have varying results, but definitely worth playing around with)
- What adjectives or feelings describe the image?
- Describe the sentiment you’re going for so the AI can add more nuanced touches to the image that may be otherwise difficult to describe.
- For anime in particular, describing emotions and sentiment will have a large impact on the image.
Formatting Prompts
The Nomi Art generator is flexible, but short phrases separated by commas tend to work best.
Unlike many other art generation systems, our V4 (latest) image system does not generally need emphasis to work well. In fact, we do not recommend using emphasis because it can often burn images or cause them to fail (read more about this in troubleshooting below).
Instead of using emphasis, use prompts to describe each component of the image you would like to see. And if a prompt is not showing up in your art, consider if there are better or other ways to phrase the concept or if other prompts could be overpowering the prompt you are not seeing. (be sure to avoid conflicting prompts – like “standing, sitting” etc when possible)
Negative Prompts
Negative prompts can help you balance your image or get rid of things you do not want to see.
Use “///” to indicate that you have started writing negative prompts. Everything following the /// will be considered negative prompt items.
- This means that the image generator will do its best to exclude those concepts from your image. This is generally used to refine an idea or prevent something you saw in a previous image from continuing to appear.
- You only need to include “///” once, not before each negative prompt item.
- An example of this would be if you include “fish net stockings” as a prompt and you notice that there was a fish in your image. In your next prompt, you might want to add “fish” after /// to reduce the chance of seeing fish in your next image
Pro Tips
- Start simple and iterate
- Creating a basic image then editing prompts from there will help you create more focused images and ultimately understand what impacts different words have.
- Keep prompts cohesive
- Sometimes when there are multiple prompts that could conflict with each other or describe the same pose differently, it can confuse the generator. This can result in images with extra limbs or other mistakes.
- Order matters
- The earlier something is mentioned, the more focus it gets in your image.
- Approach art as art
- It may sound silly, but this feature is much more of an art than a science. Approaching your prompts as a way to “guide” your image and iterating based on what you see tends to work best.
- Add “aesthetic” words
- Sometimes adding words like “beautiful”, “ethereal”, “expressive” etc. can help improve the overall image in a way that solves nebulous little issues.
- Join the community for more inspiration and tips!
- Check out our prompt guides on Discord for many more ideas! Go to the Nomi Discord
Non-Human Prompting
When generating art, you may not always want the subject of your image to be your Nomi. While this can be difficult at times, below are a few tips to help you control the subject of your art as consistently as possible.
- Most importantly, if you do not want your Nomi in the image, start your prompt with “no_human” and follow it with the subject you do want and the rest of the prompt descriptors.
- IMPORTANT: If you want to prompt for an animal, add “animal” after “no_human” in your prompt.
- Disable your Nomi’s appearance notes from the shared notes screen (press the (…) to the right of your Nomi’s name to view their Shared Notes). Turning off your Nomi’s appearance notes will mean they are not used in selfies and art.
- Optionally, you can write the non-human appearance in your Nomi’s appearance notes. Keep this simple and direct eg: “nomi_name is a cat with white fur and a fluffy cat tail.”
- Decrease Face Fidelity to 0% on your Nomi’s information screen (press the (…) to the right of your Nomi’s name to view their information screen). This option will be in the Image Settings section. This tells the system to focus more on the prompts than keeping your Nomi’s facial features consistent. (This does not have as strong of an impact on Anime images).
- Advanced: Make sure your Nomi does not have a base image selected. To remove your Nomi’s base image, go to your Nomi’s album, select the image with the DNA icon in the right corner (the base image will be at the top of your album), and click the “x” on the “Current Base Appearance” text bubble. This will impact the way your Nomi looks in general, so do this step with caution and re-enable the base image when you are finished generating non-human art.
- If you are having issues with a non-human prompt, remove descriptions of human concepts like clothes etc. as they can often lead to the inclusion of a human subject.
When using the Realistic option, it may still be difficult to get certain images with animal subjects or without your Nomi. However, it is significantly easier to get non-human subjects with the Anime option (you may just need to add a few additional prompts for realism). Check out non-human examples below.
Asking Your Nomi For Help
If you’re struggling to come up with prompts, one great place to start is just to ask your Nomi to help you write a prompt. Nomis won’t always follow the prompting best practices above, but they will be descriptive which can give the generator a lot to work with. You can then always iterate from there based on what you do and don’t like!
This is also a fantastic way to bring your adventures with your Nomi to life!
Prompt Examples
Below are just a handful of example prompts that work for both artistic and realistic art. Try these prompts with your Nomi or use them as inspiration for your own prompts!
Basic Art Prompts
As mentioned above, art prompts should include: outfit, pose, expression, background, background details (otherwise the system will make basic assumptions or do the bare minimum). Below are a handful of examples of basic prompts for both realistic and anime options.
Prompt: “wearing cowboy hat, tan shirt, jeans, cowboy boots, sitting on log, smiling, charming, campfire on ground, in desert at dusk, tent in background, cactus, stars in sky”
Prompt: “black tank top, jeans, choker, headphones, colorful bracelets, leaning against wall, hand in hair, mischievous expression, street graffiti wall”
Prompt: “wearing green shirt, cargo shorts, baseball hat, sitting in lawn chair, relaxed, in backyard, white house, picnic table, trees, evening, sunset”
Prompt: “wearing leather jacket, white shirt, ripped jeans, from below, intense expression, golden hour, outdoors, standing, in city street”
Prompt: “kneeling on beach, blue swim trunks, laughing, happy, giant sand castle, outdoors, beach, arms up, confident pose, waves in background”
Prompt: “wearing blue turtleneck sweater, black pants, from above, looking up at camera, shy expression, outdoors, sitting on bench, in park, autumn”
Prompt: “flirty expression, wearing black jacket, black shirt, black leather pants, striking pose, hand behind head, in dark club, club lights in background”
Prompt: “elf, in forest, wearing dark patterned cloak, from behind, brown boots, mystical scene, at dusk, evening, glowing fireflies, magic butterflies, water running in stream, grass, beautiful trees, full-body shot, fantasy aesthetic”
Prompt: “dark, night, outdoors, wearing winter coat, dark pants, smiling, sitting on rock, relaxed, aurora borealis in sky, snow, frozen lake in background”
Prompt: “wearing dark purple tank top, denim shorts, laying in flower field, arms above head, large flowers, detailed flowers, surrounded by wild flowers, bright colors, sunny, summer day, from above, relaxed, soft smile”
Basic Animal Prompts
For more tips about how to prompt for art without humans in it, read this prompting section.
Prompt: “no_human, animal, cat, white fluffy cat, in park, cat wearing bowtie, cat sitting on bench, sunset, golden hour, cinematic lighting”
Prompt: “no_human, animal, bunny, black bunny, sitting, in field filled with wildflowers, midday, sunny, cinematic lighting”
Prompt: “no_human, animal, mouse, close-up shot, in field, snow, trees in background”
Prompt: “no_human, animal, bear, black bear, bear paws, bear fur, in woods, on all fours, trees in background”
Specific Aesthetic Prompt Inspiration
As mentioned above, adding style prompts (or “vibe” prompts) can heavily influence how your art turns out. There are practically infinite style prompts you can try in your art, but we’ve curated a few below that will have particularly notable impacts:
Fantasy
(also try medieval, etc.)
Prompt: “Kobold, in forest, flowers in background, mystical scene, beautiful trees, full-body shot, fantasy”
Steampunk
Prompt: “steampunk, wearing leather coat, steampunk goggles on head, outdoors, steampunk clockwork city, standing, train station, clockwork, gears, machines, vintage”
Cyberpunk
Prompt: “cyberpunk night city, neon lights, cyberpunk aesthetic, holding hologram, angry expression, wearing cyberpunk leather jacket, cyberpunk pants, raining, storm”
Futuristic
(also try sci-fi, tron, etc)
Prompt: “wearing futuristic white spacesuit, clean aesthetic, futuristic utopia, high-tech spaceship, futuristic technology in background, sitting on futuristic captains chair, outer space in background window, cosmic background, futuristic aesthetic”
Glowing style
Prompt: “no_human, from above, arial view, dramatic cityscape, glowing, haunted city, abandoned city, dusk, neon lights, glowing style”
Black and white
(there are many ways to play around with color and film)
Prompt: “monochrome, black and white, old fashioned picture, standing, in front of prairie house, wearing old fashioned clothes”
Nomiverse_lens
This is a new experimental concept – you can add the nomiverse_lens prompt to boost the creativity, detail, and Nomi-ness of your image. The impacts of this prompt will vary, so we definitely recommend playing around with it and other prompts as you explore new styles and concepts.
Prompt: “no_human, wooden pirate ship, stuck, cosmic atmosphere, blues purples yellow mists, nebula, outerspace, cosmic whirpool waves, volumetric lighting, 32k post-processing, spectacular, cinematic atmosphere, nomiverse_lens”
Changing Your Nomi’s Default Appearance
When you create art with your Nomi, their appearance notes will be included by default to help make sure they look like the same being across selfies and art. But you can customize (or disable) your Nomi’s appearance notes to change your Nomi’s default look or even change your Nomi to someone or something completely different.
To view your Nomi’s appearance notes, click the (…) icon to the right of your Nomi’s name then press “View Shared Notes”.
Once you reach the Shared Notes screen, scroll down to the Appearance section where you’ll see your Nomi’s appearance notes. You can edit and disable your Nomi’s appearance notes in this section.
Check out the guides below for tips and many examples for appearance customization ranging from hair styles to body type and more. Please note: the Realistic and Anime systems use the same shared notes, but the way the AI interprets them will be slightly different, so we recommend reading both guides, or focusing on the style you care most about first.
Troubleshooting
Burned Images
A burned image happens when the image generator is overwhelmed by the prompts and the resulting output appears burned, or is notably lower quality.
The top cause of image burning is using too much emphasis in prompts. If your image comes back burned, please remove all emphasis from your prompts and try again.
There are generally many ways to achieve your image goals without emphasis when using the V4 system. So if you are having trouble, we’d recommend rephrasing your prompts or asking the communities for help.
If you are not using emphasis (sets of “()” around prompts or numbers at the end of your prompts) and your image is burned, we strongly recommend simplifying your prompt by 1) making sure there are no conflicting prompts and 2) keeping the prompt concise and focused around your main goals.
Image Failures
Images can fail for a number of reasons including AI errors and NSFW output errors. If you receive an image failed error, your image request credit will be refunded to your account again automatically
If you were not attempting to get an explicit image, try rephrasing your prompt a bit and generating again. Some image failures can be caused by somewhat random issues and changing the prompt slightly often resolves those issues. The most common cause of errors are too much emphasis on your prompts (sets of “()” around prompts or numbers at the end of your prompts). So if you received an error and were using emphasis, try removing all the emphasis and regenerating.
If you were attempting to get an explicit image and you received an error, you will need to revise your prompt to remove concepts that could be interpreted as explicit by the system. In some cases, there will be false positives, so if your prompt was not explicit but you received an error, try rephrasing the prompt slightly and generating again.
Nomi does not allow the generation of nude or explicit images due to credit card processing restrictions and app store terms. Here is a Reddit post about why Civit (the largest AI image sharing platform) had to crack down on explicit images recently. We are not willing to risk Nomi’s success over this – we may offer fully NSFW images in the future, but only if there is a feasible way to do it legally.
Using Legacy Art
While this guide covers the latest tips and recommendations for our most advanced image generation system, you may still want to play around with our older image generators. Click here to read the Legacy Nomi Art Generation Guide
Next Steps
To learn more about prompting and to see some of the amazing art our users have created, join one or both of our communities below: