I vividly remember the first time I watched Lord of the Rings. To me, it's one of the best movies of all time. I was really young when it came out, so that made the rewatching experience much better as I got older. I must have seen the entire trilogy more than ten times so far.
The reason why I'm mentioning Lord of the Rings is because it made me fall in love with the fantasy genre at a young age. Apart from that, I also played many different video games set in fantasy worlds. All of this essentially provides me with endless inspiration for creating fantasy characters and landscapes with generative AI models like Dall-E 3.
I love creating fantasy-themed images because it gives me the opportunity to work on my creativity a lot. There are so many different magical details that I can include in a prompt to ensure the generated result turns out great.
How to Write Dall-E 3 Prompts for Fantasy Images
Before I talk about how to write Dall-E 3 prompts for fantasy images, let's discuss why it's a good idea to make them at all. Apart from helping you practice your prompting skills and explore different aesthetics, you can potentially make money by creating images like this.
People who are creating video games, animated or live action TV shows/movies, comic books, visual novels, or any other form of media usually hire designers to help them create interesting characters and landscapes. You can create draft designs for small studios if you practice enough.
You can attract an audience online simply by being active on social media platforms and engaging with the community. For instance, you can share the fantasy-themed images you created with Dall-E 3 on platforms like Twitter and Instagram. Someone might discover your talent and hire you to create images for them.
In case you're a fan of the fantasy genre, it likely won't be too difficult for you to write prompts in this category. If not, you'll have to spend some time researching a fantasy glossary to learn the terminology you can use in your prompts to produce great images.
The thing that I really like about Dall-E 3 is its level of coherence. In other words this AI model is perfectly capable of understanding the descriptions in your prompts even if you only use natural language. While this is perfect for beginners, I believe that you'll get much better results if you structure your prompts in a more professional way. Let me show you exactly what I mean with some of my prompt examples that I prepared for you.
12 Dall-E 3 Prompt Examples for Fantasy Images
I have written prompts for both fantasy characters and landscapes. Of course, you can combine the two and write descriptions of both a character and a landscape in a single prompt.
You'll notice that I also instructed the model to create some images while applying a certain aspect ratio to an image. The default aspect ratio for all images created in Dall-E 3 is 1:1, but you can change it depending on your preferences by simply mentioning a specific aspect ratio in your prompts.
Now that we've got that basic info out of the way, let's check out the prompts I prepared for you today along with the images Dall-E 3 created from them.
prompt #1: fantasy character, fire-wielding phoenix sorcerer whose flaming plumage and scorching spells are fueled by the eternal cycle of rebirth
There are so many cool details featured in this image. I'm very impressed with the design of the robe. I also like the phoenix character, as well as the baby phoenix next to it. The staff also reminds me of something you'd see in a fantasy themed RPG. Let's check out what other types of characters you can create with Dall-E 3.
prompt #2: fantasy character, seasoned alchemist in a hidden mountain laboratory concocting elixirs that grant extraordinary abilities and transmuting base metals into precious ones, aspect ratio 3:4
While the first image featured the default 1:1 aspect ratio, I wanted to ensure this one had different proportions. I feel like using a portrait aspect ratio as seen in the image above is great when you're creating images where a character is the main focus.
As far as the image itself is concerned, I like everything about it. I previously mentioned that the Dall-E 3 model has an amazing level of coherence, and you can see that since it included details that show a laboratory hidden within a mountain as well as the alchemist making elixirs.
prompt #3: scene set in a fantasy world, rainbow-hued desert mirage city where structures seem to shift and shimmer in the heat, challenging travelers to discern reality from illusion, aspect ratio 16:9
I honestly had no idea what I should expect from the generated result here since the prompt includes some abstract elements. However, I was pleasantly surprised when I saw what the Dall-E 3 model created. My only issue with this image is that you can instantly tell that it's AI-generated. I prefer results that make it difficult to guess whether they were actually created with the assistance of artificial intelligence.
prompt #4: scene set in a fantasy world, ethereal astral marketplace where traders exchange dreams, memories, and emotions as currency, aspect ratio 16:9
This is so good that I wouldn't mind hanging it on my wall. The color scheme here is very impressive. I'm also proud of this prompt because it's not an idea that you'd often come across. When you have a somewhat unique idea, there's a good chance that the AI model will create something spectacular from it. Apart from the marketplace, I really enjoy looking at the background in this image.
prompt #5: fantasy character, dual-wielding rogue with shadow-infused daggers, master of stealth and subterfuge, navigating the intricate politics of a shadowy underworld, aspect ratio 3:4
The style featured here reminds me of Assassin's Creed. However, the character here comes with a much darker aesthetic. I imagine this would be a great video game character. He can be similar to the main character in Hitman, only this one is set in a world hundreds or thousands of years in the past.
prompt #6: fantasy character, druidic shape-shifter seamlessly transforming into majestic animals and harnessing the primal energies of the natural world, aspect ratio 2:3
I always liked spending time in nature, and because of that, I'm drawn to druid characters in fantasy-themed books, films, and games. This is an excellent representation of the two sides of a druid. I like how Dall-E 3 decided to create this image from the prompt I wrote. The transformation would've likely looked much different had I used some other generative AI model.
prompt #7: fantasy character, artificer engineer in a clockwork workshop constructing intricate automata infused with magical gears and arcane energy
There is something extremely special about the way this text-to-image AI tool creates illustrations. Up until the release of Dall-E 3, I hadn't seen illustrations produced by AI models look this good. And I honestly haven't seen them since. This is arguably the best AI-powered tool you can use to make illustrations.
prompt #8: scene set in a fantasy world, enchanted forest of illusions, where the very trees can cast illusions leading travelers astray or revealing hidden paths, aspect ratio 16:9
Let's go back to landscapes. I know that some people might not like prompts for characters, but they enjoy looking at these wonderful landscapes. The great thing about an image like this is that you can upscale it and use it as a wallpaper. There is something so calming about this particular image that it would make a great wallpaper.
prompt #9: scene set in a fantasy world, arcane circus traveling between realms featuring mystical performers showcasing illusions and magical feats, aspect ratio 16:9
What if the greatest show on Earth (the circus) could be even better? Imagine how great the circus would be if we lived in a fantasy world that featured magic. This is yet another image that could be used as a wallpaper. The level of detail here is absolutely mindblowing.
prompt #10: scene set in a fantasy world, labyrinth of time where pathways lead to different eras, and travelers risk becoming lost in the currents of temporal chaos, aspect ratio 16:9
Just imagine if you were in a labyrinth that could take you to different points in time. Someone should make a movie or TV show that revolves around a concept like this. I like how the model interpreted the term "labyrinth of time" here. There are many different clocks featured in this image, which definitely adds a nice touch to the labyrinth.
prompt #11: fantasy character, celestial warlock adorned in star-studded robes, wielding eldritch magic drawn from otherworldly patrons
Even though sometimes the best option is to use a portrait aspect ratio when you're creating characters, you can also get extraordinary results if you stick with the default dimensions. There are some interesting cosmic elements featured here. As always, Dall-E 3 creates yet another amazing illustration.
prompt #12: scene set in a fantasy world, hidden library in a magical forest where ancient tomes and mystical scrolls whisper secrets of forgotten spells and arcane knowledge, aspect ratio 16:9
This image looks like it's a screenshot from a fantasy video game. I wish there were hidden libraries like this in the real world where you can access arcane knowledge. But since that doesn't really seem plausible, at least I can daydream and imagine what those libraries would look like with the help of Dall-E 3.
Final Thoughts
If you've been following my website for a while, you know how much I love experimenting with AI tools. I like sharing results from most of my experiments. While writing this article, it was the very first time I attempted to make fantasy-themed images with Dall-E 3. By taking this approach, I'm able to share my genuine thoughts and impressions this AI model left on me.
What I can say overall about making fantasy images with Dall-E 3 is that this model can produce extraordinary results. I would argue that Midjourney is superior when it comes to making images that look like photographs, Dall-E 3 is much better than any other generative AI model when you want the results to include at least some form of special effects.
The images that Dall-E 3 was able to create from my relatively simple prompt examples are nothing short of amazing. The thing I like the most is that the model often picks up on even the small details you mention in your prompts. This level of coherence is currently unmatched among text-to-image AI models.
