The Image Generator is a text to image generator. You describe the picture you need, the subject, the setting, the style, the light and the framing, and the image model turns it into an original image. Your description goes to the model exactly as you write it, so the more precise the prompt, the closer the result.
It helps whenever a stock photo does not quite fit, a lifestyle scene for a product, a blog header, a background for a social post, a mood board for a client. You choose the size among those the selected model supports, usually square, portrait or landscape, and download the result.
The tool sits in the Images, audio and video family. For an overview of image creation on Infinithink, see the AI image generator feature page.
What you get
- An original image generated from your own description
- A choice of sizes based on the selected model, such as 1024x1024, 1024x1536 or 1536x1024
- Control of the style through the prompt, from product photography to watercolor or flat illustration
- Prompts accepted in any language
- The image models included in your plan, listed on the pricing page
What you fill in
- Describe the image required
-
Describe subject, setting, style, light and framing. You can write in any language; with some models English gives the most accurate results.
- Image size required
-
Available sizes depend on the selected model.
How to use it, step by step
-
Describe the subject first
Start with what must be in the picture, a ceramic coffee cup, a bright office, a mountain lake at dawn. One main subject gives a cleaner result than a list of objects.
-
Add setting, style, light and framing
Say where the subject is, the style (photo, watercolor, 3D render), the light (soft morning light) and the point of view (top view, close-up).
-
Choose the model and the size
Pick an image model available on your plan and a size that matches where the image will go, landscape for a blog header, portrait for stories.
-
Generate, refine and download
Change one thing at a time in the prompt, the light, the angle or the colors. Download the version you want and check the details before publishing.
Example, input and output
Input
- description
- Minimal product photo of a ceramic coffee cup on a light oak table, soft morning light, top view, empty space on the right for a headline
- size
- 1536x1024
Output
Illustrative example written by the editorial team.
Tips for better results
- Write the prompt as a short scene description, not a list of keywords. Image models follow sentences about subject, place and light better than isolated words.
- Name the style explicitly. Product photo, watercolor illustration and flat vector icon give very different images of the same subject.
- If the image must contain text, keep it to a few words and check the spelling.
- With some models English prompts give the most accurate results. If a detail keeps getting lost, try the description in English.
- Ask for empty space in the composition when you plan to add a headline later.
Who uses it, and for what
Blogger who needs article headers
Create a landscape header for each post in a consistent style, instead of searching stock libraries for every article.
Online shop planning a seasonal campaign
Generate backgrounds and lifestyle scenes for banners and newsletters, then combine them with real product photos.
Social media manager
Produce square and portrait visuals for posts and stories, ready to pair with captions written in the same workspace.
Agency preparing a mood board
Turn a client brief into concept images to agree on direction, light and colors before a photo shoot or a design round.
Limits and good practice
- Generated images can contain errors in small details such as hands, text, reflections or the number of objects.
- The tool cannot reproduce your exact product. On product pages, use real photos and keep generated images for scenes and backgrounds.
- Do not write prompts that imitate real people, logos or trademarks.
- Available models, sizes and image quality depend on your plan and on the model you select.
Frequently asked questions
How do I write a good prompt for an AI image?
Describe the main subject first, then the setting, the style, the light and the framing, in one or two sentences. For example, minimal product photo of a ceramic coffee cup on a light oak table, soft morning light, top view. Refine one detail at a time.
Can I write prompts in languages other than English?
Yes, you can write the description in any language. With some models, prompts in English give the most accurate results, so if the image keeps missing a detail, try the same description in English and compare. Text meant to appear inside the image can stay in its original language.
Which image sizes can I choose?
Sizes depend on the model you select. Common options are square at 1024x1024, portrait at 1024x1536 and landscape at 1536x1024. Choose the format of the place where the image will appear, so you do not have to crop it later.
Can I use the images in my business content?
You can download the images and use them within the terms of use of the service and of the image model. Avoid prompts that copy real people, brands or existing artwork, and review any image meant for advertising as carefully as any other visual.
How many images can I create?
It depends on your plan. Each image uses credits from your plan's balance, and the Free plan also has a monthly image limit. The available models and image quality can change by plan as well. The pricing page lists what every plan includes, so check it before planning a large batch.
Try it in your workspace
Start free, no credit card needed. When you need more, one suite grows with you, from Lite to Agency.
Sign in or create a free account, the tool opens in your workspace.