Generative AI has become a household tool for more than 90% of the content creators who use it to generate text, images, videos, and music. AI image generation is one of the most useful aspects of this technology.
With tools like ChatGPT and Midjourney, generating AI images is no longer arduous when you know how to craft the AI image prompts. Although writing an accurate AI prompt is no rocket science, it takes your creative & critical thinking along with the use of keywords to write a perfect AI prompt.
In this blog, we’ll take a look at how to write AI image prompts, and the best AI image prompt examples along with some of the best AI image generator tools.
What is AI Image & How Does it Work?
AI image is an image created through technologies like artificial intelligence, machine learning and neural networks. These AI systems are trained on massive datasets that learn the patterns and styles to generate unique images.
AI image generators create these images. When a user inputs the prompt that defines the patterns and themes of your images, the AI image generator uses these algorithms and datasets to generate the desired output.
However, prompts don’t merely work by short strings of words. For example, prompting “expensive car” won’t produce the desired output. You’ll need a more detailed prompt that defines the object's symmetry, shape, color, size, and other intricacies.
Thus, the AI bot doesn’t understand the word “expensive”, but it understands the features, color, size, and design of your vehicle more accurately. An effective AI prompt includes the description of features, size, color, dimensions and other artistic styles.
The neural networks use this data and browse the internet to obtain clear and concise AI images.
Read the benefits of Artificial intelligence here and what it can do for the creators!
How to Write an Effective AI Image Prompt?
Now that you have understood how AI image generators work, it’s time to get more practical and craft an AI image prompt. Here are some tips to keep in mind when crafting AI image prompts:
Know the components of your AI image
AI image prompts need to be simple and accurate. Using lengthy and complex sentences for your prompt wouldn’t accurately define your image output.
A good idea is to write your prompt, separated by commas, with a list of phrases and keywords. You can also ask AI to use color palettes and artistic styles such as realism, cubism, or expressionism.
So, whether you want to generate a realistic image of the sun on a beachside or an imaginative world of chocolates and fairies, break down the components, colors and styling for accurate results.
The Prompt Example: “Realistic image of Sun on a beachside, realistic image, outshining sun, use bright colors and joyful sunny day.”
Describe the content of your image
This is the starting point for your image generation. Describe the content of your image with as many details as you can. First off, understand your image requirements.
Know what kind of image you want to generate – Is it a sketch? A real image? Or a 3D image?
A good suggestion is to start your prompt like this:
-
A real photograph…
-
A 3D image…
-
A portrait of…
-
An illustration of…
-
A sketch of…
Describe the subject of your image
After defining the content of your image, proceed with your image subject. The subject can be an object, animal, person or an abstract thing. The image generator uses the content and subject to create a rough image, which can be enhanced further by giving a detailed description.
Examples include:
-
A 3D image of coffee table
-
A portrait of Albert Einstein
-
An illustration of Batman
Here’s Canva AI prompt - “A 3d rendering of a coffee table.”
Add more details to your prompt
Now, the AI generator isn’t an artist that uses its creativity to create art and impress you right off the bat. Instead, it needs a clear and concise understanding of your desired output.
The results can be enhanced by adding details like colors, color palettes, shapes, sizes, and textures.
Let’s say if you want to generate an image of a lion, don’t just write "lion." Add intricacies to the image by describing the type of lion (African, Asiatic, White), its surroundings (forest, desert, or a sea side), image textures and more.
Here are some examples:
-
An image of Asiatic lion with dense fur in a forest
-
An illustration of table that’s brown and shiny
-
A sketch of Commando officer in black
To enhance your image, you would further need to add the details like:
-
Define the action of your subject- For instance, a cat standing on a unicycle.
-
Mood of your subject: This tells the image generator how the subject is performing the action. For example, a happy cat standing on the unicycle.
-
Background of the image: This will define the background of your image as you need customizations to it.
Define the art form, styles and references.
Adding the form, styling and references to the image prompt will help the AI understand better so that it can scan the right datasets. As you improvise the prompt each time, the AI algorithm is trained better and comes up with a more precise result.
A variation of art forms like painting, anime, graffiti, pixels or the styles such as minimalist, surreal, impressionism can add more effects to the images. For a realistic art, giving the references would further improve the output and help AI in a more accurate result.
Let’s say, emulating an art style of “Salvador Dali”, “Andy Warhol’s Shot Marilyns painting” or “a portrait in the style of Picasso.”
“Generate an illustration of a mother holding a son, with watercolor, surreal art and Picasso like art. Use lively colors.” - generated using Canva AI
Some other examples of art forms include:
-
Sculpture
-
Collage
-
Street art
-
Textile art
-
Installation art
-
Ceramic art
Include image composition, aspects and positioning
Do not leave anything for the last minute tweaking. An AI image generator can use every detail that you add into the final output. You can use the keywords to define the image composition, resolution, aspect ratios and positioning.
Here’s how you can use the information to craft the prompts:
-
An image of Asiatic lion with dense fur in a forest with minimalist colors and expression of dominance.
-
An illustration of table that’s brown and shiny with abstract impressionism from a corner view
-
A sketch of a Commando officer in black with a wide stretched view and resolution 4096 x 2160.
In your prompts, you can also define the positioning of your subject with respect to the background.
Framing is the term that’s used to define the content or position of an image. The positioning of your subject image depends on how you start your prompt – whether you want to zoom the subject or focus on the background elements.
For instance, an image that focuses on close-ups will have a more effective prompt, with the subject mentioned first and the background after.
Here’s an example of the prompt: “Generate a realistic image of a lion with strong, fierce looks, fearlessly looking in camera, forest fire background and bright colors.”
Some AI art generators can also extend beyond this by using the prompts that define the film-shoot, camera angle and image definition. You can use "wide shot," "point-of-view," or "dutch angle” to control image composition.
Another detail to the prompt can be lightning. Adding the lightning can change the look and feel of your image. You can make use of lightning prompts such as “soft light”, “hard light”, “dynamic light” to bring depth and visual interest to the image.
Using a resolution like “4k” or “8k” will improve the details in the image and cultivate more clarity. It’s always a better idea to play around with the prompts, use a mix and match of keywords, expressions, framing and add lightning.
👉🏻Read to learn how you can use Canva AI
Other AI Prompts Tips & Tricks
While the above section is significant in crafting the AI image prompt, here are a few trips to keep in mind to mitigate errors:
-
Use the right set of keywords. Avoid using complicated keywords and phrases or something that’s not common. For AI to recognize and create, use keywords that are more readily used by internet users.
-
Use conversational tone. Since AI image generators use NLPs, the best way to wrote prompts is to use a conversational tone. This data can power the chatbot and
-
Include more adjectives. Adjectives are the prompt’s best friend. Use the adjectives to define your images as closely as possible.
-
Keep it short and sweet. Extending your prompts to more than 8 words would complicate them for the AI. A sweet spot is between 4 and 7 words to avoid being overly descriptive.
-
Avoid using terms with conflicting meanings. Using words that reflect different meanings might confuse the AI generator. For example, words like “imaginary” and “realistic” have different meanings.
-
Incorporate AI tools in the workflow. Different AI generator tools work differently. Experiment with a set of these AI generators to include it within your workflow. Examples of these AI tools are; DALL-E. Bing Image Creator and Midjourney. (more about it later)
AI Image Prompt Ideas & Examples
Here are some of the prompt examples that you can use to generate the images. Try replacing the subject while keeping the elements the same:
-
An enchanted forest bathed in moonlight, with glowing flowers and mystical creatures hidden in the shadows.
-
A clear, starry sky peeks through the dense canopy above, casting a magical glow over an ancient, winding path.
-
A bustling cyberpunk city at night, illuminated by neon signs and towering skyscrapers.
-
A sketch of a bear in bright colors and minimalist composition.
-
An image of an astronaut exploring a vibrant alien planet. In the background, a massive, ringed planet
-
Ancient samurai battling a mythological creature during a fiery sunset.
-
A photograph of a futuristic underwater city with dome lights attracting mysterious sea life.
-
A 3D rendering of an office desk with a futuristic look and bokeh.
-
An illustration of a rooftop garden in full bloom overlooking a crowded urban landscape.
-
Realistic representation of a cozy coffee shop on a rainy day, people reading and chatting.
-
A sunrise yoga session on a peaceful beach, waves gently breaking.
-
A surreal dream sequence, floating islands and waterfalls in the sky.
-
A candid photograph of a solo camper by a lakeside campfire under a starry night sky.
-
An abstract representation of the cosmic dance of planets and stars with swirls of colors and light.
-
A lo-fi representation of a secret garden at night, glowing mushrooms lighting the path.
-
An abstract painting of a rooftop garden in full bloom overlooking a crowded urban landscape.
-
A cubist painting of a candlelit dinner setup on a beach, stars reflecting on the sea.
-
A sketch of a vintage train station with steam locomotives, passengers awaiting departure.
-
A realistic picture of a small vineyard at harvest time, workers picking grapes in the golden sunlight.
-
Futuristic image of Giant robots constructing a bridge over a dazzling citylight.
-
An artist sketching the skyline from a high vantage point in the city.
-
A sketch of a rustic pottery workshop, shelves filled with clay pots and artisan tools.
-
A 3D rendering of a spaceship in the style of cubism with high resolution.
-
A breathtaking mountain landscape at sunrise, with the first light of day illuminating the peaks and valleys.
-
A lively Renaissance festival scene, with jesters, musicians, and artisans entertaining a crowd.
-
An impressionist painting of an old boot with bright violet laces.
-
Person with fearless and determined look, forest fire background, close-up shot, realistic.
-
Imaginary image of space elevator connecting Earth to a station in geostationary orbit.
-
A retro-style of robot uprising, machines and humans standing off in a devastated city.
-
A whimsical forest filled with talking animals, hidden doorways, and trees that come to life at night.
Popular AI Art Generator Tools To Try Out
There are some great AI art generators that work best for these prompts. Take a look at these AI image generators:
Midjourney
Midjourney is an AI assistant that can create images from simple text prompts. This AI art generator is best known for realistic and hyper-realistic images. In my experience, Midjourney is the best AI art generator that generates the output close to your prompts.
It also works best in resizing the image, changing the composition and editing an artistic form. However, the odd thing is there is no web app. You’ll have to access Midjourney through the Discord server.
You can see what others are creating and get inspired by their work and learn. Also, Midjourney is not free. It has a monthly subscription of $10/mo and an annual subscription of $96/year.
According to Reddit users, here are some of the most popular Midjourney references; Pulitzer Prize winning photography, fashion photography and automotive photography rolling shot.
DALL-E
DALL-E is a great AI art generator by Open AI that uses text prompts to generate surreal images instead of more realistic ones. This art generator tool interprets the text prompts better than Midjourney for creating imaginative visuals.
The interface is quiet easy to navigate and it has a low learning curve. DALL-E prompts work best in generating artistic forms like digital art, steampunk art, cyberpunk art, or vapourware art.
To fill in the background, you’ll need to mention the background type or color, else the tool leaves it white.
DALL-E is still in beta version and you can sign up to get access to it. It will give you 50 free credits when you sign up and 15 each month. If you need more, you can gain access to 460 images for $15/month.
Here’s an example of art generated by DALL-E for the prompt: “A surreal image of an eskimos celebrating near an arctic circle and show igloos, northern lights in background.”
Also Read: Best AI Tools of 2024
Bing Image Creator
Bing Image Creator is Microsoft’s answer to DALL-E. Where DALL-E focuses on surreal images & midjourney generates realistic images, Bing Image Creator is inclined more towards vibrance and details.
Although Bing Image creators need improvement in certain areas and aren't as good as the above two, adding locations, artistic styles, and camera views can optimize the image ideas better.
Here’s what Bing Image Creator generated for the same prompt: “A surreal image of an eskimos celebrating near an arctic circle and show igloos, northern lights in background.”
Can you legally use AI images?
Yes. You can use most AI images legally. In some cases, however, users should follow compliance and content policy to keep in line with usage. Some images may be protected by copyright, and hence, you need to seek permission to access them.
What is a good prompt for AI art?
A good AI prompt includes the details about the image output such as adjectives, colors, image composition, framing, textures and artistic styles and more. You can use ChatGPT to generate prompts and add your details further.
Is there a free AI image generator?
There are various free AI image generators such as Canva AI, Craiyon, Bing Image Creator, Starryai and PicsArt.
Conclusion
This brings us to the end of our comprehensive tips, tricks and the best tools to generate AI image prompts. However, these tools would be of no use until you have the right knowledge to compile the keywords, abstracts and other important words in the prompt.
Try and play with these prompts on a free AI image generator and perfect your prompt engineering to get the desirable results. It is also a good idea to reverse engineer the images to check for the prompts.
For this purpose, AI tools like NightCafe and Craiyon can be your best allies. If you completely depend on AI to generate prompts, head over to Chat GPT and Google Bard to get the right assistance.
