How to create images with AI: a guide for beginners (ChatGPT and Gemini)

How does AI image generation work?

Creating images with artificial intelligence is simpler than it seems. Describe what you want to see, in normal language, and the tool creates the image automatically. You do not need to know how to draw, use Photoshop, or have any technical knowledge. In seconds, you can have an original image created exactly for your case.

The text you write to describe the image is called a prompt. It is the same concept we use when we ask ChatGPT for text. If you are not familiar with the concept yet, read our article what is a prompt.

The main tools available

Prices and plans may change. Always confirm on the official website of each tool as of the date you read this article.

How to create an image with ChatGPT

If you already use ChatGPT, creating images is very simple:

• Open ChatGPT and start a new conversation.
• Write something like: 'Create an image of [your description].'
• ChatGPT will generate the image directly in the conversation, in a few seconds.
• If you do not like the result, ask for adjustments: 'Make the same image but with a white background and a more minimalist tone.'
• When you are happy with it, right-click to save the image.

ChatGPT uses the GPT Image model to generate images, which produces photorealistic, high-quality results. It is especially good for product images, lifestyle photography, and creative compositions. If you do not use ChatGPT yet, see our guide how to use ChatGPT.

How to create an image with Gemini

Google's Gemini also generates images in a simple, intuitive way:

• Go to gemini.google.com and sign in with your Google account.
• Write your description. Example: 'A photograph of a modern café in Lisbon, with natural light, minimalist style, and green plants.'
• Gemini generates several options to choose from. You can ask for variations or refinements.
• Click on the image you prefer to save it or share it.

One advantage of Gemini is that you can use it directly from Gmail, Google Docs, or Google Slides, without leaving the tools you already use every day. If you work a lot in the Google ecosystem, it is a very practical integration.

How to write effective prompts for images

The quality of the prompt directly determines the quality of the image. Here are the most important tips:

• Be specific about the subject: instead of 'a dog', write 'a young golden retriever running on a beach at sunset, golden sand, warm light'.
• Indicate the visual style: photorealistic, illustration, watercolor, minimalist, vintage, flat design, cartoon, etc.
• Describe the light and setting: natural light, studio, white background, cozy interior, urban outdoors, etc.
• Specify the composition: close-up, wide shot, aerial view, centered framing, rule of thirds.
• Try several versions: ask for 3 or 4 variations and pick the one closest to what you have in mind.

For more tips on writing prompts in general, visit our prompt library.

Practical uses for your business

AI-generated images can be used in many day-to-day business situations:

• Images for social media posts (Instagram, Facebook, LinkedIn)
• Illustrations for blog articles or newsletters
• Product mockups for client presentations
• Banners for email marketing campaigns
• Cover images for presentations and proposals
• Visual content for your website without needing a photographer

AI does not replace professional photography in every context, but it is a very practical and affordable solution for day-to-day needs, especially for small businesses with limited resources.

If you want to go deeper into using AI in your business in general, see our guide AI for beginners and find out how other tools can help you save time and grow.