What is DALL-E 3?
DALL-E 3 is the latest generation from OpenAI, specifically designed to create AI-generated images using artificial intelligence. With DALL-E 3 and the right prompts and examples, you can generate impressive images directly in ChatGPT through simple text inputs.
Since the latest ChatGPT4o version, creating images with DALL-E in ChatGPT is even free. This means you no longer need to subscribe to the Plus membership for DALL-E 3. Therefore, DALL-E 3 is one of the best AI tools for generating AI images at no cost.
In this article, I will introduce you to the most important features and show many examples of AI-generated images. Furthermore, you will learn how to create a good prompt for DALL-E 3, so that you make better AI-generated images by the end of this article. All images in this article were generated using DALL-E 3 in ChatGPT.
Table of contents
How does DALL-E 3 work?
With DALL-E 3, you can create very good AI-generated images even with a poor prompt. The first step is to activate the correct ChatGPT model. If you have subscribed to the Plus membership, you can select ChatGPT-4. As a paying customer, you are allowed to generate more images with DALL-E within a certain time frame.
If you are using the free version of ChatGPT, you can generate images using the ChatGPT4o version. In the following screenshot, you can see which AI models are currently available in ChatGPT.
Another alternative to use DALL-E 3 would be the Bing Image Creator. However, you need a Microsoft Account for that. I would recommend using DALL-E in ChatGPT with the free ChatGPT version. You do not need to pay any money or create an account for ChatGPT.
Additionally, with ChatGPT, you have the advantage of using Custom GPTs. For example, you can create AI images of yourself with the GPT Cartoonize Yourself.
If you are looking for an image AI to generate letters or entire sentences on an image, the new AI model Flux might be interesting for you. Compared to other AI image generators like DALL-E 3, Flux excels particularly in creating images with realistic, intact hands and correctly written texts.
For example, I created the following images using Flux in Supermachine:
Improvements over DALL-E 2
OpenAI has significantly improved DALL-E in its third version. DALL-E 3 is much more powerful than DALL-E 2. DALL-E 3 prompts are better understood, and the texts can be interpreted more accurately, which increases the quality of the AI-generated images. DALL-E 3 is better in every detail, whether it’s accuracy, colors, contrast, and more. The following images show how significant the quality difference is between DALL-E 2 and DALL-E 3. On the left side are the DALL-E 2 images, and on the right are those from DALL-E 3.
A van Gogh style painting of an American football player
3D render of a cute tropical fish in an aquarium on a dark blue background, digital art
A Formula 1 car driving on a neon road
You can see that compared to DALL-E 3 , the quality is significantly better. The prompt examples were understood much better by DALL-E 2.
DALL-E 3 Guide: Prompts & Examples
Generating Images with DALL-E 3
Generating images with DALL-E 3 is super easy. You do not need any technical knowledge or anything similar. Of course, it is beneficial if you are familiar with prompting techniques. However, it is not a necessity.
You can already generate images with a minimal DALL-E prompt like “Create an image of a Labrador.” However, you can achieve better results with advanced DALL-E prompts. For example, you can improve the quality of the images with the following prompt:
Create an image of [main subject], that is [description of activity or pose].
The image should take place in [environment or context], where [specific environmental details].
The main subject is wearing [description of clothing or accessories] and has [specific physical features or expression].
The atmosphere is [mood or atmospheric conditions], with [lighting conditions or time of day].
The style of the image should be [desired artistic style], with a focus on [specific stylistic or technical details].
You can adjust the aspect ratio by appending “Square aspect ratio”, “Widescreen aspect ratio”, or “Rectangular aspect ratio” to your prompt.
So far, I have only had good experiences with this technique, and it is by far the best prompt template for DALL-E 3.
DALL-E 3 Prompts: Examples and Inspiration
Create an image of a young artist painting on a large, colorful canvas. The image should take place in a brightly lit studio, where sketches and paint tubes are scattered everywhere. The artist is wearing a smudged apron and has curly, tousled hair. The atmosphere is creative and inspiring, with natural light streaming through large windows. The style of the image should be impressionistic, with a focus on vibrant colors and dynamic brush strokes.
A futuristic city where houses stand on giant mushrooms.
A steampunk universe where airships hover over steam-powered cities and mechanical dragons dominate the sky.
A futuristic metropolis where robots and cyborgs live side by side with anthropomorphic animals like rabbit detectives and fox engineers.
A young woman with long curls, emerald green eyes, and a charming smile.
Advantages and Disadvantages
Advantages
- Free in ChatGPT with the ChatGPT4o version
- DALL-E 3 in ChatGPT is intuitively designed
- DALL-E 3 can generate images in various styles and formats
- The images are generated within seconds
Disadvantages
- No adult content allowed, and strict censorship
- Limited generation of images with the ChatGPT4o model
- It is always recognizable that photorealistic images are AI-generated
- Realistic DALL-E 3 prompts typically result in cartoon-style images rather than photorealistic ones
In my opinion, the advantages outweigh the disadvantages, especially since using DALL-E 3 with the ChatGPT4o model is free.
Supermachine is an alternative if you are looking for an AI image generator for 18+ content. Click here for my Supermachine review.
DALL-E 3 vs. Midjourney
The most well-known AI image generators include DALL-E and Midjourney. Both the DALL-E AI model and Midjourney have been trained on millions of different images. The comparison between the two models is not easy. DALL-E is probably the simplest way to generate images using artificial intelligence. DALL-E can interpret your prompts better and usually optimizes them.
Image Quality and Customization
Recently, DALL-E 3 has introduced the ability to customize AI-generated images using inpainting. However, Midjourney has the edge here, as making adjustments in Midjourney is much easier, and creating consistent characters is also easy for beginners.
Costs and Commercial Use
DALL-E 3 can be used for free in ChatGPT. No account or subscription is required – as long as you use the ChatGPT4o version. Midjourney does not have a free version, and the cheapest Midjourney plans start at $10 per month.
The commercial use of AI-generated images is possible with both AI image generators. Of course, images created based on existing protected brands do not count. For example, you will not have image rights to an AI-generated image created based on the Nike logo or similar.
DALL-E 3 Alternatives
DALL-E 3 primarily competes with Midjourney and any AI image generator based on Stable Diffusion, DALL-E, or Flux. Examples include the AI image generators Supermachine and Artsmart.
The best alternative is Automatic1111 and a Stable Diffusion model from Civitai. With suitable hardware, you can create AI images for free and without limits.
Frequently Asked Questions (FAQs)
Is DALL-E 3 free?
Since OpenAI made the ChatGPT4o model available, DALL-E 3 is free. You do not need an account or the Plus membership to create AI-generated images.
How much does an image cost on DALL-E?
The costs depend on the quality and resolution. For example, an image with standard quality and a resolution of 1024×1024 costs $0.040. Further prices for using DALL-E and the other AI models can be found in the pricing overview from OpenAI.
Can DALL-E edit its own images?
DALL-E 3 supports an inpainting function. You can mark an area with a brush and then regenerate it using a prompt from DALL-E. In this video, I show you how to edit images in DALL-E 3 in ChatGPT.
