If you follow AI artists on social media platforms, you've likely noticed that many of them often post anime style images that they made with the help of text-to-image AI tools. That's because tools like Stable Diffusion and Midjourney do an excellent job of creating high-quality images in this art style.
I will be using the preset 'Anime' style option available in Stable Diffusion XL to make all of the images in this article. I'll also divide the prompt examples into two categories - anime characters and anime landscapes. There will be nine prompt examples for each category.
I personally use the Clipdrop platform made by Stability AI to get access to Stable Diffusion XL 1.0, which at the time of writing this article is the latest Stable Diffusion model. If you want to run this AI tool locally, you can do that. I have a simple guide where I explain how to install Stable Diffusion locally. Check it out if you're not sure how to do it.
How to Write Stable Diffusion Prompts for Anime
Regardless of whether or not you're a fan of anime content, you can still make beautiful images in this art style. Since there is a preset style option in Stable Diffusion XL that's designed specifically for anime style images, the process is incredibly simple.
The first piece of advice I'd give you if you're not too experienced with AI text-to-image tools is to keep your prompts relatively short. You don't want to make them more complicated than they should be. If your prompts are too long, you might risk confusing the model and getting an unexpected result.
The second thing I'd say is to include elements from fantasy, sci-fi, and other genres where fictional realms are filled with creatures, magic, and natural phenomena that don't exist in the real world. Doing this will help make your images more interesting and alluring.
My third and final piece of advice before we go through the prompt examples I wrote and prepared for you today is to let your imagination run wild. You have the complete freedom to give any instructions you want to the AI model. Use this freedom to let your creativity flow. That's honestly one of the best benefits of text-to-image tools.
Since I've started using tools like Midjourney and Stable Diffusion, I've noticed that my imagination has become more vivid. I feel like a kid again whenever I decide to make images with the help of AI. These tools have truly had a positive impact on my overall mood.
18 Stable Diffusion Prompt Examples for Anime
Let me start off this section by mentioning that I have a full article on how to make anime images in Midjourney. If you're a fan of the Midjourney text-to-image AI model, I suggest you check out that article to check out some amazing prompts for anime and the results generated from them.
As I already mentioned in the intro to this article, I will divide this section of the article into two different categories. The first category will be all about anime characters. Meanwhile, the second category will consist of images that depict landscapes in an anime art style.
9 Stable Diffusion Prompt Examples for Anime Characters
I'll be using a 4:5 aspect ratio for every anime character prompt example. There is an option in Stable Diffusion that enables you to select this aspect ratio. Now, let me start showing you the examples I wrote for you today.
prompt #1: character covered in a lush, living tapestry of flora, with vines and leaves that appear as a natural, enchanted armor
Everything in this image is marvelous, from the background to the armor. This can serve as excellent concept art when developing a character that you'd want to use in a comic book or animated content. As you can see, you don't need a complex prompt to generate a good and unique image. All you have to do is clearly describe what you want the AI model to create.
prompt #2: character with a third eye in the center of their forehead, which grants them the ability to perceive hidden truths and glimpses of the future
I didn't want the third eye to be a literal eye, so I was happy when I came across this image among the generated results. This is a type of character you'd see in a fantasy-themed anime. The background here really matches the third eye well.
prompt #3: rogue with a mischievous grin and a set of daggers, lurking in the shadows of a bustling city
Wow, this character looks dangerous. I love the aesthetic of the image, and how the main character is in full focus. You can't see other people in the image well but that doesn't matter because your main focus should be on this rogue.
prompt #4: character whose skin seems to be made of molten silver, with flowing liquid-metal limbs that shimmer and flow like quicksilver
When you're creating fictional characters in Stable Diffusion, you should focus on describing their appearance in the most concise way you can. Try to think outside of the box and come up with concepts that you think would look aesthetically pleasing.
If you want to check out an in-depth guide on how to create characters in Stable Diffusion (that are not in an anime art style), I have an article dedicated to that topic. I suggest you check it out if you're looking for useful information and want to see some beautiful prompt examples.
prompt #5: character with a cloak made of iridescent feathers, which shimmer with every movement and create an ethereal bird-like aura
I had to generate a huge amount of images to get one that I liked. When you think you wrote a good prompt, you shouldn't give up until you see an image that you think is worthy of using.
prompt #6: character with wings of ice dwelling in a frozen tundra
Alright, this is an incredibly simple prompt, but sometimes that's all you need to make a great image. Of course, you also need a bit of luck for the AI model to create a stunning result. The matching of colors in this image is superb. The blue eyes and blue details in the armor go perfectly with the ice background.
prompt #7: character with emerald-green eyes and a crown of nature, controlling the elements to protect a lush, enchanted forest
I guess that the use of the adjective 'emerald-green' made the AI model interpret it as that there should be emerald stones in the image. I don't mind this one bit, as I think it adds a nice touch.
prompt #8: elderly sage with a long, flowing beard, holding a weathered staff and surrounded by ancient runes
Now this is a nice concept art image. The hands aren't perfect, but that's okay. At the time of writing this article, Midjourney v5 does a much better job of generating hands than the Stable Diffusion XL 1.0 model. I really love the background here, it gives off an aura of mysticism and divine knowledge.
prompt #9: character with a mechanical arm and an assortment of gadgets, operating as a genius inventor in a steampunk-inspired laboratory
I've already mentioned this in a few articles, but I find the steampunk art style quite beautiful. When you combine it with an anime style, you often get incredible results. Some small details seem off in this image, but it's overall a great result with many intricate details.
9 Stable Diffusion Prompt Examples for Anime Landscapes
It's now time to change the aspect ratio that you'll use. Since you're now making landscapes, you can choose between the 3:2 or 16:9 aspect ratio. Even though the 3:2 aspect ratio option in Stable Diffusion is literally called 'Landscape', I prefer to use the 16:9 aspect ratio.
prompt #10: luminous, bioluminescent forest at twilight, with radiant trees and glowing flora that create an otherworldly, ethereal ambiance
This is an image I'd gladly use as a wallpaper. But I recently made a post about how to make high-quality wallpapers in Stable Diffusion and created one that I'm going to be using for a very long time. The glowing flora definitely adds a nice touch to the forest, which creates a truly ethereal ambiance.
prompt #11: tranquil, bamboo forest with sunlight filtering through the dense green canopy, casting a soothing light on the winding paths
What a nice, calming landscape. Although the canopy is not that dense, I don't really search for perfection when I'm creating images in Stable Diffusion or any other text-to-image model. The prompts I write represent more of a general idea of what I want the AI to create, and then I'm happy to sift through the results until I find an image that's aesthetically pleasing.
prompt #12: ice cavern deep within a glacier, with walls adorned in shades of blue and turquoise, illuminated by the eerie light of glacial ice
This would be a great hideout. I've been to an actual ice cave before and it was a tremendous experience. There are nice shades of blue in this image. It looks like something that would be in the opening scene of an episode of an anime show.
prompt #13: cascading waterfall hidden within a dense, enchanted forest, surrounded by vibrant oversized mushrooms
It's interesting to see how a simple idea can have a big impact on a generated image. In this case, the idea was to include oversized mushrooms in the landscape. I feel like they're what make the image so great. It could've been just a regular landscape in an anime style, but the mushrooms make it seem like a fantasy realm.
prompt #14: hidden valley nestled between towering cliffs, where a crystal-clear river winds its way through lush, emerald-green meadows
This looks like a village where a hero lives. I didn't expect to see houses in this landscape, but they are a nice touch. There's not much to say about this landscape other than that it's beautifully designed.
prompt #15: ancient desert with towering sand dunes, bathed in the soft, warm light of a setting sun
This looks like an area that the main character of an anime would have to walk through to reach a certain destination. I didn't expect to see this much vegetation in the desert, but that's okay. It can be edited out if need be.
prompt #16: serene, misty moorland at dawn, with ancient stone circles and standing stones shrouded in fog
I would've been very surprised if I actually saw stone circles similar to Stonehenge in the generated images. Although the chances of the AI model generating something like this were extremely slim, I decided to include it in the prompt anyway. But I also included the 'standing stones' keyword just in case. The result is great, with many different colors blending together as well as huge monolithic stones creating a stunning landscape.
prompt #17: icy wasteland beneath a star-studded night sky, aurora borealis
The glowing ice structure in the middle of the image is absolutely breathtaking. It looks like it absorbed the aurora borealis.
prompt #18: twilight cityscape overlooking a pristine, calm ocean, with sparkling skyscrapers and a bridge that spans the tranquil waters
Just look at the orange, pink, and blue hues in the water. That alone made this image stand out for me among the numerous generated results I looked through.
Final Thoughts
Hopefully, you picked up a few tips on how to create anime style images in Stable Diffusion from my prompt examples. I specifically wanted to make a distinction between characters and landscapes because I know that some people prefer one over the other.
As you can see, there is nothing too difficult about writing prompts like this. The only thing you have to do is use your imagination to construct scenes. And that may be easier said than done in the beginning, but once you make it a habit to write prompts often you get used to it pretty quickly.
If you're new to generative AI models, I encourage you to take some of my prompts and make small changes to them to see how they affect the end result. This will teach you a lot about how these tools work.
