One of the earliest text-to-image tools is by far the best we’ve tested. With Dall-E, OpenAI pioneered the way for generative AI that turns text prompts into images. The competition is stiffer now, but version 3 of the service still holds up.
In tests comparing it to Adobe Firefly and Google ImageFX, I found Dall-E 3 to work best with realistic and engaging images, and almost always best with surreal illusions. It’s a bit slow, but it’s most likely to give you good, usable results on the first try, especially if you’re looking for interesting, not failed AI illusions.
Dall-E 3 is available only through the premium ChatGPT Plus service, which costs $20 per month and also gives you access to a more responsive version of the ChatGPT chatbot and OpenAI’s helpful GPT Store, which contains customized versions of its AI tools. If you want to get a feel for what’s possible, you can try the earlier Dall-E 2 for free, but it’s not as good.
OpenAI says it may use content submitted to Dall-E 3 to improve the model’s performance, that it shares content with a select group of “trusted service providers,” and that it doesn’t sell data or share content with third parties for marketing. You can also submit a privacy request to have OpenAI stop using your data for training or delete your account. For more information, see OpenAI’s general privacy FAQ and main privacy policy.
Here are my further findings about Dall-E 3.
How CNET tests AI image generators
CNET takes a practical approach to reviewing AI image generators. Our goal is to determine how good it is relative to the competition and what uses it’s best suited for. To do this, we give the AI prompts based on real-world use cases, such as rendering in a specific style, combining elements into a single image, and handling long descriptions. We rate the image generators on a 10-point scale that takes into account factors like how well the image matches the prompt, how creative the result is, and how responsive it is. Check out how we test AI to learn more.
How is the quality of the image? How well does it match the prompt?
ChatGPT is the best of the text-to-image AI tools I’ve tried at producing useful, interesting, and believable results. It still makes plenty of mistakes, like a pickleball player’s bat growing out of his head instead of the bat grip, but the results make me want to explore further instead of closing the browser tab. It does a much better job with dynamic scenes, contact and interaction between different subjects, and emotions.
ChatGPT is a big part of Dall-E. It amplifies your prompts, adds ornate text, and injects drama into the results. It also fosters a conversational style of use: You can request an image, then request an adjustment, without having to resubmit the entire query.
ChatGPT’s language technology also enables it to handle long, complex prompts. Advanced word processing capabilities, it turns out, are very useful for advanced image processing capabilities.
This helps Dall-E 3 outperform competitors including Adobe’s Firefly and Google’s ImageFX in translating your prompts into what you want and assembling multiple elements correctly. For example, Dall-E 3 was the only AI image generator I tested that successfully created a dragon flying over a castle, breathing fire, and holding a white, furry sheep in its claws. Sure, it held the sheep tenderly, perhaps in response to OpenAI’s draconian rules against violence, but it’s close.
Don’t expect perfection in AI-rendered images of exploding balloons. Dall-E made a lot of detail mistakes. In an image of a dog walker overwhelmed by too many dogs, the human looked ridiculous surrounded by dozens of them. But if you look closely, you’ll see the usual AI issues: One of the dogs has two heads, another is a cat, and the others have disturbing problems with their legs, ears, and tongues. But the image is still striking.
Perhaps we’ll collectively develop a new set of aesthetic rules to accommodate AI images, accepting a few errors in exchange for their value. If so, Dall-E 3 will do a great job. How attractive are these images?
Very attractive. Dall-E 3 produces vivid, eye-catching images again and again. Even with the issues, I often like them. They sometimes make me laugh and observe the details.