How to use the AI DALL-E 3 in ChatGPT from OpenAI?

by | Feb 20, 2024 | Tutorials | 0 comments

Whether it’s for specific needs, to boost your SEO, or simply to have images that look just like you, artificial intelligence is the ideal answer. With DALL-E 3, OpenAI gives this trend a major boost, combining the performance of its generator with the ease of use of ChatGPT. It’s as if image creation became accessible to everyone. In this article, we’ll tell you all about its performance, explain how to use it, and tell you all about its limitations.

DALL-E 3, the easy-to-use image generator

all images man smiling cameraOffered free of charge by Microsoft under the name “Image Creator”, or available to ChatGPT Plus subscribers, DALL-E 3 is a powerful, easy-to-use tool. There’s nothing superfluous: just ask it with the same ease as a question in ChatGPT. You give your instructions in the conversation thread and get illustrations generated by artificial intelligence. Queries, known as prompts, become more effective the clearer and more precise they are. You can refine the generated image by adding details in the dialog box thread. For example, you can specify the setting in which your character will evolve, or adjust the lighting to suit your preferences.

Choose your style, from painter to photographer

DALL-E 3 opens up a whole range of illustration options: painting, graphics, photography, comics, you name it. You can even specify a particular artistic style: cubism, realism, pointillism, steampunk or pop art. And if you’re lost, all you have to do is mention a famous artist, and the AI will immediately understand. Ask for an Andy Warhol-style portrait of a woman, and the result will surprise you. You also have the freedom to adjust lighting effects, such as corner lighting to create a dramatic effect. For photography enthusiasts, the demand for a silver or photo booth style generates breathtaking effects.

Create characters and their worlds

Looking to create the image of a character to illustrate a story? Using simple, explicit French, explain your idea. For example, ask for the image of a woman in her fifties, with gray hair and an obvious elegance. Specify the framing, whether it’s a portrait, a close-up or a wide shot, and even adapt the background scenery to your liking. It may sometimes be necessary to formulate several requests to arrive at the precise image you have in mind. Invented characters are also easy to create, so you can let your imagination run wild with a cat-headed mouse or an alligator in a three-piece suit.

Then put them on stage

all images group candles tableWith DALL-E 3, you can design an exceptional backdrop or use an existing one to integrate your character. This feature lets you create images that are as realistic as they are bizarre and unexpected. Imagine your elegant fifty-something lady moving from Times Square to a science-fiction universe. The potential is truly infinite, enough to awaken the creative spirit within you.

The limits of DALL-E 3

Even if your imagination knows no bounds, DALL-E 3 has its own. You can’t ask for images of contemporary celebrities such as politicians or athletes. On the other hand, ask Albert Einstein to play tennis in black and white – it works! And beware: the AI categorically refuses images of a pornographic or sexual nature, as well as violent images. Microsoft is adamant on this point: circumventing the rules can lead to account suspension. And for image tinkerers, DALL-E 3 doesn’t allow you to modify an existing image, unlike other AIs.

Tips for mastering DALL-E 3

DALL-E 3 is a revolution, combining text and image generation in a single tool (ChatGPT). But to achieve your goals, you need to know how to handle the “prompt”. It may be necessary to explore different approaches, perhaps even making more than ten attempts, before finding satisfaction. So don’t hesitate to diversify your language to stimulate the AI, by favoring the use of English in your queries as much as possible. Indeed, the AI seems more at ease with English vocabulary and the results are more consistent. Vary your vocabulary and lexical fields. Be precise in your indications, impose your style, your choice of colors, the layout of elements, lighting, texture, even movement. In short, position yourself as the conductor of your own creation, and create your own unique and personal work.

Discover our photos for free!

We don't want you to leave... Did you know that you can download our images for free and without registration?