My AI assistant has also created a podcast episode for the following article. If you prefer listening to reading, you can listen to the podcast via the following link (ATTENTION: Podcast is created exclusively by AI, no guarantee for accuracy).

AI technologies such as ChatGPT can not only generate texts, but also create images for us. There are also specific AI tools, such as Midjourney, DALL-E or Leonardo.ai, which focus exclusively on creating images. In this article, I’ll show you how to create unique images with an AI using the right prompts.

---

To repeat

What are AI tools?

AI technologies, also known as AI tools or apps, are programs that use artificial intelligence (AI) methods to offer their users added value. A well-known example of AI technologies is ChatGPT. The AI tool has become much more popular over the last two years, primarily due to its high level of user-friendliness and the fact that it is free to use. This type of AI tool uses a special type of artificial intelligence – generative artificial intelligence. Thanks to the use of generative AI, modern AI technologies are able to produce new content that did not previously exist in this form. AI technologies support their users in a completely new way and are capable of scaling more and more human processes. ChatGPT is very generic and not very focused on individual use cases. However, there are many other generative AI technologies that focus on very specific applications such as podcast creation, meeting notes processing, research, image creation and many more. A commented overview of tested AI technologies for relevant applications can be found in the Generative AI technologies overview.

---

What AI technologies are available for images?

Midjourney

Midjourney is currently probably the most popular image AI tool on the market. Until now, however, the tool was considered a professional tool without a free version. However, since the end of August 2024, all users with a Googel account can use Midjourney Web (free of charge)*. With the link: midjourney.com/imagine, anyone can create a free account and generate AI images. However, you should pay particular attention to the image rights.

*The free version was unfortunately only available for just under 5 days, now there is also a small subscription at Midjourney.

DALL-E

DALL-E is the image AI from Open AI. Currently, users can only use DALL-E in GPT 4 and there is no free version.

For good results, describe the mood to create the right feeling. This also includes describing the lighting conditions or the time of day. In addition, users can also ask DALL-E to exclude things from your pictures or not to add any independent details. If groups of people are to be depicted, tell DALL-E how many, otherwise the bot will tend to squeeze as many people as possible into one picture.

DALL-E has the disadvantage that it outputs square images by default. Determining other aspect ratios in the prompt only works moderately.

Stable diffusion

https://stablediffusionweb.com/#google_vignetteStable Diffusion is an open source tool from Stability AI. The Free Plan allows users to generate two images at the same time. With the Premium subscription, you always have a choice of four images. These are stored for seven days. Users can also upload images and then have them converted.

Canva Magic Media

In the Pro version of Canva, users can also generate images with an AI. Canva also relies on Stable Diffusion. The advantage here is that you can generate four images at the same time. Canva also has DALL-E integration. Users can therefore use both AIs and see which one works better.

Leonardo

Leonardo enables art, images and videos to be created with the help of AI. The tool helps to realize creative visions and visualize everything from anime to photorealistic portraits and 3D textures. A wide range of AI tools are offered and the platform uses a variety of models that are specially adapted to different styles and requirements, giving you a high degree of flexibility when generating images. The main features of Leonardo include: Art and image generation: With simple text instructions you are able to create anything from detailed illustrations to complex graphics. Video animation: Static images can be easily converted into stunning animations to tell dynamic stories or enhance your presentations. Transparent PNG creation: You can also create images with a transparent background for web design and product presentations. 3D texturing: You have the option of uploading 3D models and Leonardo will create suitable, realistic textures for you.

Using AI tools for image generation: Step-by-step guide

Image generation with AI tools is more than just a simple prompt. As with the use of other AI technologies, we must first create our own concept for AI image generation or consider what we actually want.

Preparation

First you have to develop your own idea. What kind of picture do you actually want? What is the image used for? Who is the target group of the picture? What do you want to say with the picture?

The more precise your own ideas for the image are, the better the prompt for the AI image generation and your final result will be.

Implementation

When you have a “picture of your picture in your head” you can start with the prompting. Describe the image you want to create in short and precise sentences. Be specific and detailed when describing the image and use a concrete structure and language. (You will find specific prompt instructions later in this article).

Improvement

In most cases, it doesn’t work on the first attempt. So make use of the feedback and improvement options and give the AI further prompts until the AI image meets your expectations.

Utilization

Finally, all you have to do is download the AI-generated images and then you can use them. But pay attention to data protection information and be transparent. Label your AI-generated images as such. You can do this with a subtitle or a watermark, for example.

3 important tips for AI prompts for images

I have 3 simple tips for beginners in the field of AI image generation. If you take these into account in your prompt, you will get the first good results.

1. Description Subject: What do you see?
2. Details and surroundings: What about it?
3. Style, artist, media type: What does it look like?

How do I set up a prompt for AI image generation?

If the 3 tips above are not enough for you, then you are already an advanced user. In general, the prompt should then be structured as follows:

• Image type (photo, logo, style…)
• Main motif (landscape, person, animal, object…)
• Scenery (surroundings…)
• When (time, light…)
• How (artist, camera, color…)
• Other parameters and styles

Professional tips for AI images: Even more details

If the prompt tips shown above are still not enough, you can add the following categories to your prompt. However, it should be noted that AI images are often created by non-professionals. Not every AI user is familiar with photography or image generation in the deeper sense. The following professional tips therefore require a sound background knowledge of art, images and photography.

• Art styles: Abstract, Abstract Expressionism, Academism, American Realism, Anime, Art deco, Art Nouveau, Arts and Crafts, Atompunk, Baroque, Bauhaus, Biopunk, Classical Realism, Clockpunk, Conceptual Art, Cubism, Cybernoir, Cyberpunk, Dark Fantasy, Decopunk, Dieselpunk, Digital Art, Expressionism, Fantasy Realism, Flowerpunk, Fine Art, Forestpunk, Futurism, Gothic, Harlem Renaissance, High Fantasy, Impressionism, Installation Art, Manga, Modern Art, Modernism, Neoclassicism, Neo-Impressionism, New Realism, Op Art, Photorealism, Pixel Art, Pop Art, Post-Impressionism, Postmodernism, Precision Art, Realism, Rococo, Romanticism, Socialist Realism, Steampunk, Surrealism, Synthwave
• Painting Types: Acrylic Paint, Airbrush, Canvas, Cave Painting, Chinese Painting, Coffee Paint, Color Field Painting, Dripping Paint, Fine Art, Glass Painting, Gouache, Graffiti, Hard Edge Painting, Hydrodip, Wall Painting, Oil on Canvas, Oil Paint, Painting, Paper Marbling, Puffy Paint, Rock Art, Scroll Painting, Splatter Paint, Spray Paint, Still Life, Street Art, Tempera Paint, Tibetan Painting, Watercolor, Wet Paint
• Print styles: Advertising, aquatint, banner, barcode, block print, blueprint, brochure, business card, collage, coloring book, comic, cyanotype, election photo, election poster, etching, graphic novel, halftone, illuminated manuscript, illustrated brochure, instruction booklet, intaglio, linocut, lithograph, logo, magazine, “Magic the Gathering” card, manuscript, map, mezzotint, monoprint, film poster, newspaper, newspaper print, photo collage, photography, stamp, poster, product photo, propaganda poster, QR code, scheme, signage, silver gelatin, sticker, storyboard, storybook illustration, tarot card, ukiyo-e, visual novel, wall sticker, woodcut
• Adjectives: strange, ancient, angelic, angry, fearful, athletic, award-winning, simple, beautiful, messy, cheerful, clean, cold, colorful, confusing, cozy, creepy, cute, depressing, detailed, dirty, disgusting, dreamy, dry, ecstatic, older, ethereal, evil, excited, expensive, fancy, fat, flat, flat design, flat shading, fluffy, friendly, furry, fuzzy, gloomy, good, adorable, creepy, hairy, happy, very detailed, huge, hyperrealistic, impossible, incoherent, complicated, intricate maximalist, joyful, big, lonely, clear, luminous, massive, massive scale, mature, gentle, micro, mini, minimalist, moody, morbid, mottled, muted, nano, nervous, OCD, old, ornate, otherworldly, photorealistic, plain, powerful, pretty, priceless, psychedelic, calm, rainy, realistic, refreshing, sad, simple, eerie, sleepy, smooth, spooky, strong, surface detail
• Lighting: accent lighting, afternoon, artificial lighting, backlighting, beautiful lighting, blue hour, bright lighting, lit by candlelight, Christmas lighting, cinematic lighting, colored lighting, backlighting, twilight, dark lighting, dawn, daylight, daytime, subdued lighting, dramatic lighting, twilight, evening, film noir lighting, lit by firelight, flickering light, floodlight, fluorescent light, front lighting, global lighting, golden hour, semi-dark lighting, halogen light,
• Time periods: Ancient Egypt, Ancient Greece, Ancient Rome, Antiquity, Assyrian Empire, Aztec, Babylonian Empire, Benin Kingdom, Bronze Age, Byzantine Empire, Carolingian Empire, Dark Ages, Edwardian Age, Elizabethan Age, Georgian Age, Gilded Age, Great Depression, Heian Period, Inca, Industrial Revolution, Iron Age, Maori, Mayan, Medieval, Meiji Period, Mid-Century, Middle Ages, Ming Dynasty, Minoan, Modern, Moorish, Mughal Era, Nasrid, Navajo, Neolithic, Olmec, Ottoman Empire, Paleolithic, Persian Empire, Pre-Columbian, Prehistoric, Qing Dynasty, Regency, Renaissance, Retro, Shang Dynasty, Songhai, Stone Age, Sumerian, Tokugawa Shogunate, Tudor, Victorian, Viking, World War I, World War II, Zhou Dynasty, Zuni Pueblo, 1100s, etc.
• Decorative arts: 3D printing, amigurumi, applique, balloon modeling, balloon twisting, bas-relief, beadwork, blown glass, bone china, carved, carved ivory, carved lacquer, carving, kneading, cloisonne, crochet, cross-stitch, diorama, embroidery, enameling, felting, fretwork, glass mosaic, ice carving, impressionist mosaic, marquetry, inlay, puzzle, crochet, lacquer, lampwork, lath art, leather carving, leatherwork, marble, marquetry, micromosaic, miniature painting, modular origami, mosaic, needlework, origami, paper model, paper cutting, papier-mâché, photographic mosaic, pietra dura, porcelain, pottery, doll, puzzle, pysanky, quiltwork, quilting, relief carving, repousse, origami, sand art, scrimshaw, sculpture, stained glass, statue, string art, tapestry, tattoo, tattoo art, Venetian glass, weaving, wet folding, carving, wood burning
• Rendering techniques: 3D Model, 3ds Max, 500px, Arnold Render, ArtStation, Blender Render, CGsociety, Cinema4D Render, CryEngine, Cycles Render, Daz 3D, DeviantArt, DirectX Render, Doughy Render, Houdini Render, Infini-D Render, KitBash3D, Luxcore Render, Marvelous Designer, MentalRay Render, OctaneRender, Optix Render, Photobashed, Photoshop, physically based render, Pixia, Quixel Megascans, Raylectron Render, Redshift Render, Sketchfab, Substance 3D, Terragen, Unreal Engine, Vray Render, Weta Digital, Zbrush Render
• Photography Styles: Daguerreotype, Tintype, Film negative, Tri-X, Kodachrome, Slide film, Portra 800, Natura 1600, Ilford Delta 3200, Polaroid, Hasselblad, Double exposure, Multiple exposure, Large format camera, Wide angle lens, Fisheye lens, Tilt shift lens, Anamorphic, Lensbaby, Telephoto lens, Prime lens, f1.8, f2.8, f4, f11, f16, photoshoot, commercial, thermography, X-ray, infrared
• Artists: William Logsdail, Beatrix Potter, Roy Lichtenstein, Richard Corben, Michaelangelo, Gerhard Richter, Bjarke Ingels, John Berkey, George Inness, Peter Andrew Jones, J.M.W. Turner, Todd McFarlane, Caravaggio, Atey Ghailan, Hirohiko Araki, Huang Guangjian, Ray Caesar, Takeshi Obata, Antoine Blanchard, Diego Velázquez, Romero Britto, Guido Borelli da Caluso, Lucas Cranach the Elder, Nele Zirnite, Bob Ross, Zdzislaw Beksinski, Glen Fabry, Jane Graverol, Krenz Cushart
• Colors: black, silver, grey, white, maroon, red, purple, fuchsia, green, lime, olive, yellow, navy blue, blue, aquamarine, aquamarine
• Common expressions that can enhance results: Masterpiece, trend on artstation, trend on pixiv, vivid, dynamic, geometric, intricate, high quality, detailed

What do I need to bear in mind when using AI images?

Apart from the right prompts, there are a few other aspects that you should consider when generating AI images.

Image rights

If you have images created by an AI technology, this does not mean that you automatically have all the rights to them. It depends on what can be seen in the pictures. If logos are shown, for example, this can be problematic.

A distinction must also be made between rights of use and copyrights. To really be on the safe side, you should inform yourself comprehensively and consider seeking legal advice if necessary.

Uncanny Valley

As already mentioned, there are a few pitfalls to watch out for with pictures of people. But even if you have eliminated these and achieved an anatomically correct result, there is sometimes something unreal about AI images. You can’t put your finger on it, but something is wrong. This effect is known as the “Uncanny Valley”. A well-known example is the animated film Polar Express, in which the characters in the film were not accepted by the audience. Animated films therefore often feature animals or other non-human beings to avoid the effect. Another method is animation styles that are less photorealistic.

So think carefully about whether you really want to have humans represented by AI.

Trial and error

It sounds so simple: I describe in a few words what the picture should look like and in a few seconds I get a result that corresponds exactly to my ideas. No endless back and forth, no feedback loops. Unfortunately, the reality is different.

AI tools can deliver solid results, but often they won’t be what you actually wanted. A designer who knows your brand and with whom you have worked before can realize your ideas much more precisely and understand your vision better than an AI can. They can also respond to your feedback individually and understand your change requests better.

The devil is in the detail

In addition to extra fingers or uneven eyes, there are often less obvious errors that creep into AI-generated images. Especially when many objects are depicted, mistakes can quickly go unnoticed. If you then publish the picture without checking it again and the mistake is noticed, it looks unprofessional. That’s why you should always take a very close look.

Any more questions?

Then write me a message with your wishes and questions and we will find an offer for you. Just send me a message via WhatsApp or email.

Or come directly to my WhatsApp group – where I regularly post use cases, news, best practices, events and much more about chatbots, ChatGPT and co.

By the way, I have also created a general prompt guide. It can be downloaded directly from the following page.

This article is also available as a podcast episode

Attention! The podcast was created entirely by my AI-Assistant based on my contribution – no guarantee for incorrect content.