AI Image Generation: How It Works and How to Use It
AI image generation creates or transforms pictures from a text prompt or an existing image. The main modes are text-to-image (make a new image from words), image-to-image (restyle or transform a photo), and expand/uncrop (extend an image beyond its edges). This guide explains each and when to use it.
What AI image generation is
A model turns your input — a text prompt, a reference image, or both — into a new image. You stay in control through the prompt, the reference, and the aspect ratio, and iterate until it fits your use.
Text to image: from prompt to picture
Type what you want and generate. Text-to-image is ideal for concepts, backgrounds, marketing visuals, and anything you do not already have a photo of. Pick an aspect ratio that matches where the image will be used.
→ Open text to imageWriting prompts that work
Be specific about subject, style, lighting, and framing, and add what you do not want. Short vague prompts give generic results; concrete details ("soft daylight, top-down, on a marble surface") steer the model. Iterate a few times rather than expecting one perfect shot.
Image to image: restyle & transform
Start from an existing photo and change its style, background, or look while keeping the subject. This is how a plain product shot becomes a studio or lifestyle image, and how you keep a consistent style across a set.
→ Open image to imageExpand or uncrop an image
Need a wider crop or a different aspect ratio than the original? Uncrop (outpainting) extends the image beyond its edges, generating matching surroundings — useful for reframing one image for feed, story, and banner sizes.
→ Open uncropAspect ratios for social & e-commerce
Generate at the ratio you will publish: 1:1 or 4:5 for feed, 9:16 for stories and short video covers, 16:9 for banners and thumbnails. Producing at the target ratio beats cropping later and avoids awkward compositions.
Using AI images in ads & marketing
AI images work for ad creatives, thumbnails, social posts, blog headers, and A/B test variants — produce several angles cheaply and keep the ones that perform. On paid plans the outputs come with commercial rights.
Try it yourself
A practical guide to AI image generation: text-to-image, prompting, image-to-image restyling, expanding (uncrop), aspect ratios, and using AI images in ads and marketing.
Try text to imageFAQ
What can AI image generation create?
New images from text, restyled versions of your photos, expanded crops, and edited images — for products, ads, social, and concepts.
Can I use the images commercially?
Yes — outputs on paid plans include commercial rights for marketing and product use.
What formats and quality do I get?
Standard web image formats at resolutions suitable for listings, ads, and social; generate at a large size and downscale per channel.