Let me tell you a little secret! Writing Dall-E prompts is comparatively easier than writing prompts that give you text-based answers.
But to get images that are worth using, you must know a few technicalities and the way you must structure Dall-E prompts.
In this article, we have brought prompts for Dall-E to you that experts have used and gotten great results.
What are the Best Dall-E 3 Prompts?
The best Dall-E-prompts are the ones that exhibit clarity, specificity, and creativity while effectively guiding the model to produce images that exceed expectations.
We’ve curated a list of the best Dall-E-3 prompts using which you can create images that look like they’re designed by a pro designer.
1. Design Mockups for Clients
Mockups can save you a lot of time and money in the design process. Especially for the client-side misunderstanding, mockups help present a raw design to the client and identify potential problems in the design beforehand.
Recently, while finding Dall-E-3 prompts from experts on X (earlier Twitter), we came across Joseph Hillenbrand, a graphics designer living and working in the Twin Cities.
Hillenbrand uses Dall-E capabilities to create blank mockups for business cards, posters, billboards, and packing boxes. And he’s been doing it efficiently using pin-pointed prompts.
Here are two of his intriguing mockup designs using Dall-E 👇
Prompt 1: Blank poster on an exterior wall in New York City, with a person walking by the poster in long exposure, photorealistic.
In another post in this thread, Hillenbrand uses Dall-E to create an image of a packaging box. You can use these images as mockups by laying over your brand logo and other essentials.
Also, you see how Hillenbrand uses ‘Photoreal’ or ‘Photorealistic’ at the end of his prompts to give a style statement to AI — eventually making the outputs more realistic and natural.
Prompt 2: Blank white box sitting on a white sheet with harsh lighting, photoreal.
2. Prompt for Product Designs
One of the things we noticed is that DALL-E is great at designing products for things that don’t exist yet. This is one of the things the AI excels at creatively.
M Vijaya Kumar shares a prompt that he used to ideate and create a handbag design. Carrying all the characteristics of a good prompt, it’s clear, specific, and finely detailed. 👇
Prompt: Photograph of a handbag with a pattern of snowflakes and reindeer design bold dark brown and charcoal color palette, minimalistic background.
3. Generate Tattoo Ideas
Dall-E-3 is being used and experimented to its full potential. Tattoo artists are also taking advantage of the generative AI model to ideate impressive tattoo designs from their imagination.
A Reddit user, also a Tattoo artist, shares how Dall-E-3 helped him come up with designs for a tattoo idea.
Prompt: An arm tattoo of a psychedelic cartoon-style dog with a mischievous grin, smoking a cigarette and wearing a hat. The text “GOOD DOG, BAD HABITS” is written above the dog in a bold, handwritten font.
4. Best Dall-E-3 Prompt for logo design
Oftentimes, graphic designers struggle to ideate logo designs for clients. Well! Not anymore.
Here’s an X post by Janelle Shane from a few years back. She used Dall-E to redesign the logo of Dr. Pepper, a popular soft drink, in many variations and styles.
Prompt: Logos that are dark red, black, and white, like the real Dr. Pepper logo. All have bottles or cans incorporated into the logos, which are spelled like “Derpper” and “Depper” and “Pepeper.”
5. Design Creative Signboards
Dall-E-3 excels in putting life to your ideas by understanding your intent better than the earlier versions.
NationYell (Reddit username) shares how Dall-E-3 can design signboards for various use cases. In this case, it was for a protest.
The user intended to create something that could grab attention and hurt no feelings while spreading the message loudly and clearly.
Eventually, the prompt also conveyed a particular style, context, and background, allowing the generative AI model to create an accurate image.
Prompt: “Anarchy > Capitalism” on a sign carried in a protest in the style of 1970s Newspaper Comics.
6. Make a Die-Cut Sticker Design
An AI educator, Chase Lean, took to X to share DALL-E 3’s impressive sticker designing skills.
Lean used a pillar prompt specifying the object and the number of stickers. Every time he wanted a pack of stickers for a new object, he simply replaced ‘objects’ with names such as Cactus, chicken vs. ninja cats, desserts, Christmas tree, and more.
Here’s the main prompt template he used to generate stickers:
Prompt: 9 different stickers featuring [objects] with vibrant colors and white borders on a minimal background.
👀 Note: If you want the border to be cut around the edge of the design instead, add the words “die cut” to the prompt.
7. Dall-E Prompt for 3D Illustrations
Eric from OpenAI shares a Dall-E-3 prompt that’ll help you generate detail-oriented, high-definition 3D illustrations with absolute precision.
There are many things about his prompt that we love, for example:
- Using ‘extremely detailed’ at the end of the prompt acts as an art modifier. It adds more precise details to the output instead of simple art, but it can also make the art overwhelming.
- Add an image quality modifier — a ‘100mm’ lens. Extremely zoomed-in photo with two-thirds of the background blur and looks like somehow photographed from a far distance and then zoomed in a lot.
Prompt: Claymation art of an old man playing guitar, 100mm, candle lighting, industrial colors, extremely detailed.
8. Design Hand-Drawn Illustrations
Dall-E-3 can replicate the nuances of a hand-made sketch or drawing. You can use it to create images in diverse illustration styles and monochromes.
Tanner from OpenAI tested out Dall-E for hand-drawn illustrations, asking the AI model to generate ball-point pen art.
Prompt: Cat wearing sunglasses sitting in a chair, ball-point pen art.
Dall-E’s response was spot on! 👇
However, do you see how Tanner specifies the ‘illustration style’ in the prompt?
Likewise, Dall-E can work on a range of art styles, including:
- Stencil, street art
- Pencil sketch
- Charcoal sketch
- Woodcut
- Line art
- Etching
And many more.
For instance, another Reddit user created this stunning pencil sketch using Dall-E.
Prompt: A pencil sketch of a marshmallow playing hockey.
9. Generate Book Cover Illustrations
As of 2023, there were over 200 e-books in Amazon’s Kindle store that say ChatGPT is a writer or co-writer. No wonder publishers have started using ChatGPT to write books, especially children’s books. Now, with Dall-E-3, you can also illustrate book covers within seconds.
Ammaar Reshi, head of design at Elevenlabs.io, took to X to share his experience with Dall-E-3 to generate book cover illustrations.
Prompt: Illustration of a book cover with hand-drawn art. A group of children sits in a circle in the jungle, surrounded by animals like toucans, tigers, and frogs. The atmosphere is lively and filled with the sounds and colors of the jungle. The title ‘COUNTING CRITTERS’ is written playfully at the bottom.
10. Comic Book Art
Gone are the days when designing comic art would take days or weeks. Today, Dall-E-3 does the work within minutes.
Nathan from OpenAI shares a prompt that helped him create an exact replica of a comic book scene.
The good part about his prompt is that it’s pin-pointed, specifies the art style ‘comic book art,’ and guides the AI with clear instructions about the scene.
Although the text in the picture isn’t clear, the image is accurate.
Prompt: The Wii Fit Trainer presents the Incredible Hulk with divorce papers, comic book art.
Another Dall-E-3 user, Ammaar Reshi, uses the generative AI model to recreate a Batman comic scene. However, this time, unlike Nathan’s prompt, Ammaar uses details and modifiers, like:
- Camera angle: Close up
- Colors: Black and white
- Dialogue placements: Over his (Joker’s) head
And as expected, the results are far better than Nathan’s prompt. Check it out. 👇
Prompt: create a black and white comic book panel with a close-up of the Joker towering over Robin lying on the ground. Include a comic bubble over his head saying, “Bats can’t save you!”
11. Dall-E Prompts for Infographic Designs
Let’s be honest; infographics are complex to design. They incur too much detailing and require designers to fit a million things into an 1800 x 600-pixel box.
However, Dall-E-3 boasts the capability to design infographics seamlessly. Bryan Marley took to X to share a comprehensive 20-prompt guide for Dall-E-3 users to elevate their design game.
Among those, here’s his prompt for creating high-quality infographics using AI.
👀 Note: We have added some size and style modifiers to this prompt for more accurate results.
Prompt: Create a [size] pixels instructional style infographic about [specific topic].
12. Dall-E-3 Prompt for Product Packaging
Over 72% of American consumers say that product packaging design influences their buying decisions.
A designer would take weeks or even months to come up with different variations and styles for packing your product. But what if Dall-E could help ideate some concepts beforehand?
We have come up with an exact prompt for creating product packaging using Dall-E-3. Check this out. 👇
Prompt: Generate packaging ideas for a new [product type], and name it [product name], use miniature elements through the packaging to cover blank spaces.
What are the top tips to create better Dall-E prompts?
If you want to go that extra mile with Dall-E and create designs like a pro designer, here are a few tips for getting the most out of Dall-E prompts:
1. Use Prompt Modifiers for Accurate Outputs
Prompt modifiers are like keywords or terms added to prompts that guide AI in generating accurate images. They highly influence various aspects of the generated images, such as subject, style, and quality.
A study on prompt modifiers by Jonas Oppenlaender divides modifiers into six types:
- Subject term: Include words that describe the subject and context of the image to AI. For example – an old car in the garage.
- Art style modifier: Instruct AI to create images in a defined artistic style like Oil painting, digital art, as a cartoon character, or as spray painted on a wall.
- Quality boosters: Words added to increase aesthetic qualities and the level of detail in images. Extremely detailed, award-winning, Photorealistic, Fanart, etc.
- Image Quality: Elaborate on the pixels and frame you want to use for the image. For example, 4k/8k, 15mm wide-angle lens, Close up, Macro.
- Repetitions: Repeat words or terms to enhance their value to Dall-E, resulting in better outputs. For example, a very very very beautiful car.
- Magic Terms: Terms and phrases that inject randomness and creativity, like control the soul.
Similarly, there are many other ways to use words and terms as modifiers. A Reddit user shared a comprehensive guide for prompt modifiers, perfectly categorized with images (and we loved it).
2. Use Descriptive Adjectives
Descriptive adjectives help refine the prompt for the AI model to understand it for specificity better. Use them to elaborate on the image you desire from Dall-E-3.
For example:
Prompt without adjectives: A sunset sky over the sea.
Prompt with descriptive adjectives: A fiery red sky over a calm blue sea.
Prompt without adjectives: A lady’s handbag.
Promit with descriptive adjectives: A medium-sized lady’s handbag, designer, red-colored, with two bite-sized deers engraved on it.
You see, context is key when working with DALL-E 3. The more detailed and specific you are with the prompts, the more accurate, precise, and high-quality images you’ll get from AI.
3. Try Multiple Variations
Although Dall-E-3 works on a powerful text-to-image model, it sometimes fails to grasp user intent from basic prompts.
The problem is that most users quit on Dall-E-3 after one failed attempt to generate the desired image. What they don’t know is that AI is much better at correcting itself via follow-up prompts.
Karol Roszak, a generative AI specialist, shares how he used follow-up prompts to iterate the first image that Dall-E-3 generated into multiple beautiful variations.
At first, Roszak asked AI to generate an oil painting.
Prompt 1: An oil painting of a winter snowy landscape…high quality, visible brush strokes.
Next, Roszak prompted AI with a follow-up, asking for some iterations in the generated image.
Prompt 2: Alright, now change it to raw professional photography, and the northern lights should be green now.
Prompt 3: I want it to be in 2D animation style.
Can you see how easily Roszak iterates the generated image without re-entering the whole prompt every time?
Similarly, instead of reacting negatively to an undesired output, try asking for variations of it using custom follow-up prompts.
Dall-E vs. Midjourney: Which one should you use?
Although Dall-E-3 and Midjourney share similar techniques to generate images, both have their way of interpreting prompts and rendering images.
This means that images generated using Dall-E-3 vs Midjourney are utterly different in style, quality, and intent.
Check out the table below to compare the nuances of using Dall-E and Midjourney and find out when it makes more sense to use one over the other.
| Dall-E-3 | Midjourney | |
| Accessibility | Easily accessible through ChatGPT, the Bing Image Creator, Microsoft Paint, and other services using API. | It doesn’t have a web app. Users have to access it through Discord. |
| Ease-of-use | The conversational interface is easy for people with no design expertise. | Has a learning curve attached to it and might be confusing for first-time users. |
| Power and control | Very limited control and lacks customization options except for prompt-based iterations. | Best-in-class prompting with built-in editing tools to customize outputs on the go. |
| Cost | $20/month | Starts at $10/month |
| AI Image copyrights | Users own the generated images. | Users own the generated images. |
| Integrations | It integrates with Zapier, Slack, and other daily-use apps, fostering automation and inclusiveness in daily processes. | Has fewer integrations than Dall-E-3. |
The choice between Dall-E-3 and Midjourney should be relatively straightforward for marketers:
- If you want to edit images on the go and you’re happy spending time learning through the process, Midjourney is a great choice.
- If you want an easy-to-use AI image generator that provides you with high-quality images without requiring any prerequisites or skills, Dall-E-3 is what you need.
Collaborate with your team on Dall-E
Generating images with Dall-E-3 is fun, but you know what’s more exciting? Doing it together with your team!
Team-GPT lets you onboard your team and have conversations with ChatGPT together.
To generate images with Dall-E-3 via Team-GPT, all you have to do is create an account, click on the ‘Generated Images’ button from the left menu, and write the prompt.
Iliya Valchanov
Iliya teaches 1.4M students on the topics of AI, data science, and machine learning. He is a serial entrepreneur, who has co-founded Team-GPT, 3veta, and 365 Data Science. Iliya’s latest project, Team-GPT is helping companies like Maersk, EY, Charles Schwab, Johns Hopkins University, Yale University, Columbia University adopt AI in the most private and secure way.