Learn why OpenAI’s latest AI image generator might be one of the best values in AI image generation. The recent update from Dall-E-2 to the latest generation Dall-E-3 model makes it one of the top tier image generators and allows users to conversationally interact with an image to adjust various details for the next iteration.
Who is OpenAI?
OpenAI is a leading AI research entity that develops advanced AI models most notably ChatGPT for natural language processing and DALL-E for image generation from text. Founded in 2015 with a commitment to open collaboration and sharing.
While many people are familiar with ChatGPT, they can be less familiar with DALL-E their image generation model now in its third generation. OpenAI seemed to have an early lead in image generation and then it took a back seat to competitors like Midjourney and Stable Diffusion while focusing on their flagship product ChatGPT.
Luckily, the new version of DALL-E is as impressive for images as ChatGPT is for text and programming.
Let’s take a look!
What is DALL-E-3?
OpenAI DALL-E-3 is the latest version of AI image generator released by OpenAI. It’s a massive improvement over the previous DALL-E-2 model released in July 2022. To highlight how far it is come in a little over a year the following example uses the same prompt under both models.
Example of DALL-E-2 Image Generation (2022 AI Model)
Prompt: cinematic shot of a whaling boat off of the coast of New England during a storm in the style of Dunkirk directed by Christopher Nolan –ar 2:1
Example of Dall-E-3 Image Generation (2023 AI Model)
Prompt: cinematic shot of a whaling boat off of the coast of New England during a storm in the style of Dunkirk directed by Christopher Nolan –ar 2:1
The resolution of these images have been downscaled down for better presentation on a website, but as you can see the results are literally night and day different. DALL-E-3 is capable of creating stunning art and life like images based on text prompts.
DALL-E-3 also takes text prompting to a new level compared to some of its competitors.
AI Image Generation Meets ChatGPT Conversations
The images that DALL-E-3 are stunning, and at least on par with some of the other advanced AI image generation tools on the market today. The true game changing aspect of this product is the integration with ChatGPT-4. If you’ve used ChatGPT in the past or an AI language model you’ve probably used it in a conversational way. Ask a question, get an answer, and ask a follow-up question relative to the previous response.
Now you can do the same thing with DALL-E-3 and image generation.
The examples below show DALL-E-3 being accessed from ChatGPT Plus, OpenAI’s paid tier of ChatGPT that costs $20 (USD) per month per user using the following prompt.
Prompt: cinematic shot of a whaling boat off of the coast of New England during a storm in the style of Dunkirk directed by Christopher Nolan –ar 2:1
Upon receiving the response, we tested out a follow-up question asking to make the image darker, grittier with the shadow of a whale in the water. ChatGPT returns two adjusted images based on the original renderings.
Conversational AI Image Generation Example
As you can see, the images are adjusted based on the text prompt. The interpretation of our request isn’t quite what we were looking for, but it shows a lot of interesting promise. Our expectation is that the chat style interface of adjusting images will improve over time as OpenAI collects more data to determine what people are actually looking for when requesting specific changes.
With the new twist on user interface, ability to ask questions, and ability to generate stunning AI images, we highly recommend looking at DALL-E-3 when considering signing up for an AI image generation subscription. It’s an incredible value if you’re already using ChatGPT and only need to generate images on occasion.
Now let’s look at pricing.
OpenAI DALL-E-3 Pricing Explained
There are two ways to access OpenAI DALL-E-3. Most retail customers will gain access by signing up for ChatGPT Plus which includes access to DALL-E-3, ChatGPT-4, and advanced capabilities such as the ability to search Bing with AI, and upload PDF files for AI analysis. ChatGPT Plus costs $20 (USD) per user per month. The second option is to access DALL-E-3 through an API on a per use basis DALL-E-3 API Pricing.
DALL-E-3 Usage and Resolution Limitations
DALL-E-3 is built on top of the ChatGPT-4 model and is subject to the same usage limitations of 50 messages every 3 hours. This should be enough image generation capacity for most small businesses and casual users. People looking to generate larger quantities of images may be better served with services such as Midjourney that offer premium plans with unlimited quantities and faster image generations while still producing high quality images.
Images generated with DALL-E-3 are also limited to a maximum resolution of 1024 x 1024. We hope this is updated in the near future as it limits the ability to deploy this at scale for professional graphics designers, and advertising use cases.
DALL-E-3 Image Licensing
The general consensus regarding AI images is that they cannot be copyrighted. However, the images are generated based on training datasets of other images and AI image generation is such a new field that final legal answers have not been fully determined at the time of writing and are subject to change.
Even without concerns of copyrighting, there are a lot of AI image generators that forbid commercial use of images generated by their products for commercial use as part of their terms of service. OpenAI DALL-E-3 allows for commercial use along with a number of other options available on the market today.
ChatGPT Plus is One of the Best Values in AI
At $20 (USD) per user per month, we find that ChatGPT Plus offers a lot of value for businesses and individuals. Its definitely worth consideration if you want to subscribe to a cutting-edge AI platform that acts as a general language model tool to answer questions, help with computer programming, or as a digital assistant and have the capability to generate amazing AI images within the same platform.
ChatGPT Plus also includes a number of additional features as noted below.
The features include:
- Default – A traditional language model, but faster with processing priority and better answers than most.
- Browse with Bing – Enter a search term, ChatGPT will search Bing, and summarize the answers from results across the internet.
- Advanced Data Analysis – The ability to upload files for analysis, or perform data analysis operations using Python, the world’s most common computer programming language.
- Plugins – 3rd party add-ons that allow ChatGPT to interact with other cloud services such as Expedia and Instacrt.
- DALL-E-3 – Conversational interface for advanced AI Image generation.
We recommend people check out these different features. While $20 (USD) seems like it can be a lot it can also do a lot more than it’s competitors within a single interface. ChatGPT was a breakout success and OpenAI continues to improve the original model to keep it ahead of its rivals.
Anybody who seriously uses AI will find enough value in just the LLM aspects of the service not including everything else it can do.
Conclusion
OpenAI DALL-E-3 is a huge leap forward from their prior DALL-E-2 model. It brings OpenAI’s image generation solution up to par with more advanced competitors and adds a unique twist where you can request changes to images in a conversational way using the ChatGPT-4 model.
The usage limits might be a deal breaker for people looking to generate images at resolutions greater than 1024 x 1024. For most casual users and people that are mostly interested in the use of ChatGPT as an LLM to get work done faster with the occasional use of an image generator it can be an exceptional value and at least worth consideration for an all-in-one AI experience.