Skip to main content
All CollectionsHelp Guides
How to Generate Images with Leonardo.Ai
How to Generate Images with Leonardo.Ai

Get started with AI image generation in a few simple steps

M
Written by Marc Santos
Updated over a week ago

Welcome to Leonardo.Ai! We’re glad you’re here. This guide will get you started in creating your first images on our AI-first creativity suite. Create visual assets from illustrations to marketing materials with unprecedented quality, speed, and style-consistency.

 After logging on to Leonardo.Ai, head over to our web application and follow the following steps.

1. Navigate to AI Image Generation

Click on Image Generation in the Get Started Here section in the top, or in the AI Tools in the left sidebar.

 The Image Generation page will have preloaded settings including a Preset, Preset Style and settings like the Generation Mode, Image Dimensions and Number of Images.

 As you become familiar with Leonardo.Ai, you'll eventually learn to adjust these settings according to your use case. For generating your first image, let's keep the default settings.

Presets are predefined combinations of generative models and Styles alongside other additional settings that are meant to allow users to easily create images in specific styles effortlessly. You may consider them as an overarching genre type to an certain style of imagery. There are presets for cinematic, anime and illustrative type images and more.

Preset Style: These allow you to generate images in highly specific styles. For example within the Cinematic Kino preset, you may opt for the HDR or Moody style which will give you two distinctively different photorealistic aesthetics.

Generation Modes: There are two generation modes to Leonardo's Image Generation tool. Fast and Quality. Below is a quick rundown of the two choices.

​ Fast - Using Fast mode allows you to generate images much quicker than Quality mode. However this comes with an reduction in quality and resolution. You may notice that faces that are mid-distance and further away will not look as good in this mode.

​ Quality - With this mode, you are able to generate higher quality images at higher resolutions at the expense of generation time. This also allows for better refinement of details that may otherwise not be as good when using Fast mode.

2. Enter a Text Prompt

In the prompt bar under the Image Generation header, type in a text prompt describing what you would like to see, for example, an orange cat by the window sill.

 Note that unlike language models, image generation works best with descriptions of what you would like to see with the usage of keywords. That is, instead of:

"Please draw me an orange cat by the window still."

 A better text prompt would be:

"an orange cat by the window still"

 An typical prompt would include 3 parts, Subject (eg. a woman or a cat), Medium (eg. analog photography or paint canvas) and Style (eg. long exposure photography or impasto). Being as descriptive as possible with your prompt goes a long way in getting the image you want, but it is best to start with simpler prompts.

 With Presets and Preset Style, you may choose to opt for not entering any Style or Medium into the prompt and picking from the available options instead.

⚠️ Note: In image generation, text prompts are descriptions of scene, not an instruction to the AI model.

3. Click on the Generate Button

After typing in your text prompt, click on the Generate button. This will start the image generation process which will usually take around 30-40 seconds with Quality mode selected. Note that the button will state the token cost of the generation job.


 Images will be displayed on screen once generation is finished. From there, you may download the image and use it for your project.


In just a few steps, you've generated a high-quality image on Leonardo.Ai.

As you may have noticed, one of the key ingredients to making great images is the text prompt. In the next guide, you will learn how to leverage Leonardo.Ai's features to Write Great Text Prompts.

Did this answer your question?