Leonardo.ai is an AI-powered art generator that can produce beautiful images effortlessly. However, to create stunning images using leonardo.ai, it’s crucial to master certain techniques. This guide will offer you tips to help enhance the quality of your images.

Envision your goal and look for inspiration

Before attempting to create an image in Leonardo.ai, It’s important to have a clear idea of what you want your image to look like. This will help you use the tips in this guide effectively. Try to be as specific as possible, thinking about the image’s style, setting, location, and other details.

You can find inspiration by exploring Leonardo AI public archive of images made by other users. If you see an image you like, you can remix it.

With remixing you can use the same prompt, model, and other options that the original creator used when creating an image.

Use Leonardo AI presets

One of the best things about Leonardo AI is that it offers a variety of settings to optimize your design. However, choosing the right options can be tricky. Thankfully, Leonardo Image Gen v2 has presets that make it easier by helping you start your creation with the best settings tailored to your design intent.

Before you generate an image, start by selecting the right preset for your intended design. Leonardo will then pick the best settings, including the model, to match your Image needs.

leonardo ai presets in image gen ui — Location of presets in Leonardo ai image gen v2

Experiment with models

Even though selecting a preset for your design can often choose the right model for the image you’re generating, sometimes it might not give you the results you want. Don’t be afraid to experiment with different models to understand what each one produces. Here are some Leonardo AI model suggestions you can try:

For vibrant, colorful images or images with excellent lighting, try using the Leonardo Diffusion XL model.
For photorealistic images, consider starting with the Leonardo Lightning XL model or Leonardo Vision XL. These models can create images with natural lighting and usually don’t create high-saturation images like the Leonardo Diffusion XL model.
If you are looking for cinematic-looking images a good starting point is the Albedobase XL, and Kino model.
To create anime-style photos and digital illustrations, the Leonardo Anime XL Model and the 3D Animation model are excellent choices.

If the default Leonardo models don’t provide what you’re looking for, you can also try out community models or even train your own custom model.

Improve images by writing better prompts

To create better looking images in Leonardo, it is important to understand proper prompt engineering. This section provides tips to improve your Leonardo.ai prompts.

a) Prompt writing

Writing prompts that creates beautiful images requires consideration of a few key factors:

Prompt execution order

Prompts are processed from left to right, meaning words at the beginning carry more weight than those at the end. To generate better images, begin your prompt with the primary subject. This ensures the subject has a significant influence on the resulting image.

Let’s say you want to generate a queen wearing a red dress in front of a burning castle. but you want the burning castle to be dominant in the image. you can try this prompt:

“burning castle at night time, ultra-real queen wearing a red dress”

burning castle and queen wearing redress image by leonardo ai — Image credit: leonardo ai

if you want the queen to be dominant in the image, you can try this prompt:

“ultra-real queen wearing a red dress, burning castle at night time”

queen wearing a red dress by leonardo ai — Image credit: Leonardo ai

In the first image, the castle and fire were the dominant elements. However, placing the queen at the beginning of the prompt generated the second image where the queen is the dominant subject.

Subject Context

What separates amateur and skilled AI artists is the consistency of their results. Beginners may occasionally produce good images by chance, while skilled artists consistently generate unique, high-quality images. The secret to their success is the level of detail they include in their prompts. Elements such as lighting, location, coloring, and action provide valuable context for the AI.

To improve your ai generated images, add as much context to the prompt as possible. let’s say you want to create an image of an ape king. and you write the prompt “ultra-detailed ape king sitting on the throne”, you might get a simple image :

ape sitting on the throne image by leonardo.ai — image credit: Leonardo.ai

Modified prompt – “ultra-detailed ape king sitting on the throne, focus on the subject, hard cinematic lighting, high vibrance.”

high quality image of a king ape generated by leonardo ai — Image credit Leonardo AI

The modified prompt included more details such as lighting, focus, and vibrance level. The generated image was now more visually appealing than the previous image without enabling alchemy.

When adding context, it is important to avoid making the prompt excessively long, as the model may struggle to accurately interpret it. Start with a few words and gradually add more details as you refine your prompt.

b) Let Leonardo AI suggest prompts.

Sometimes crafting a great prompt can be challenging and you can run out of ideas, but thankfully Leonardo comes with helpful features that can help you with your prompt:

1. improve prompt feature: use this feature to add more descriptive details to your basic prompt.

improve prompt in leonardo ai image gen v2 — Improve prompt feature in Leonardo ai image gen v2

2. prompt generation feature: This tool generates prompts based on a concept you provide. For instance, if you input “futuristic man,” Leonardo will create a series of prompts that align with this theme. Use this feature when you’re seeking fresh inspiration or exploring new creative directions.

c) Negative prompts

Let’s face it, AI image generators can sometimes include unwanted details in generated images like nudity, extra fingers, and more. However, Leonardo.ai’s negative prompts feature can help you remove unwanted features from your image.

Negative prompt is a feature that allows you to specify the elements that you want to exclude from the generated image, and Leonardo will attempt to remove them.

Take advantage of Prompt Magic

Prompt Magic is a powerful tool available on the Legacy mode of Leonardo AI image gen that can refine your AI-generated images. It enhances details, lighting, and adherence to prompts, allowing you to create photorealistic images or achieve specific artistic styles such as anime, impressionism, or hyper-realism.

Experiment with prompt magic strength

When using Prompt Magic, you need to experiment with different strength settings to achieve your desired image. There is no ‘best’ setting for this feature, and I recommend you try different strengths to see what works best for the image you are creating.

Low strength (below 0.3) enhances your image a little while considering your prompt. However, this setting may lead to significant deviations from your desired outcome.
High Strength (Above 0.4) At this level, the AI might strictly follow your prompt. This might make the pictures it creates less creative.

Prompt magic v2 vs Prompt magic v3

Prompt Magic v2: is available to free users, this version improves prompt adherence and composition, but it gives you less control.
Prompt Magic v3 (PMv3): Exclusive to premium users, this version includes a Raw Mode option which can work with more detailed prompts. PMv3 simplifies the generation of various types of images, eliminating the need for complex prompt engineering.

Improve images with magic words

To distinguish your image creations on Leonardo.ai, consider including ‘magic words’ in your prompts. These simple and powerful words can significantly enhance and add detail to your images.

1. Clarity magic words

These words influence the sharpness and definition of your image.

sharp edges
ultra-real
highly detailed
crisp
focus on subject
smooth texture

2. Lighting magic words

Using these words can drastically change the light and colors in your picture.

hard light
phosphorescence lighting
Electroluminescence lighting
studio lighting
photo luminescence
bioluminescence
neon
dramatic lighting

3. Color magic words

These words can adjust the color saturation and balance in your image

vibrant
natural colors
high saturation
colorful
color balance
muted colors
vivid colors

Depending on the style of the image you’re creating, you can incorporate one or more magic words from each group into your prompts. Experiment with different combinations to discover what works best for your image.

If you want to see bigger changes, especially when using color and lighting magic words, try using the Leonardo Diffusion XL model.

Copy beautiful images

You don’t always have to start from scratch to create a great-looking image. With the help of Leonardo AI Image Guidance, you can use reference images to create new images that are influenced by a reference image.

You can adjust the influence of the reference image and ControlNet options to select the ControlNet type. below is an image showing a reference image used to generate a new image using image to image ControlNet option.

Ai generated image using monalisa as reference — Image generated in Leonardo using Monalisa as a reference

The latest version of Image Guidance comes with new ControlNet options, such as Style Reference and Content Reference features. However, these features are only available to premium users. They allow you to generate new images that use styles or elements from other images, making it easy to create good-looking images by providing other good-looking images as references.

Reference Styles in Prompts

Guide the AI tool by specifying the style you want for your generated images. you can do that by including distinct styles or referencing famous artists in your prompt.

If you’re creating a cartoon image or illustration, you can reference popular cartoon and anime styles.

Consider this example: Let’s say you want to generate an anime-style image of a female superhero, most people would write their prompt like this “anime female superhero, superhero pose” and they might get a simple-looking image. This is what I got using that prompt with Leonardo Diffusion XL model selected:

anime female superhero generated by leonardo.ai — image credit : Leonardo AI

I wasn’t pleased with the results so I decided to reference Hirohiko Araki’s style used in JoJo’s Bizarre Adventure anime. I modified the prompt to “Hirohiko Araki art style, anime female superhero, superhero pose”. The modified prompt produced a beautiful image:

better looking image by leonardo ai referencing Hirohiko Araki style — Image credit: Leonardo AI

Boost Your AI Image Creations with Leonardo Elements

Leonardo elements are like photo effects that can add a unique touch to your creations. They’re simple to use and guide the AI’s output towards specific styles. By adding Elements, you gain control over the image’s style, making your creations unique.

leonardo ai elements — Leonardo ai elements selection section

To use an Element, just click on Add Elements and pick the style you want. If you are on Leonardo AI image gen v2 UI, you can add elements by clicking the image icon on the left side of the prompt input form and scrolling to the elements section in the popup.

After selecting an element, I recommend adjusting its weight. Start with a modest value around 0.2 and increase as needed.

If you’re using multiple elements, set a higher weight for the element you want to have the greatest influence on your creation. Begin with a weight of around 0.3 to 0.4 for the dominant element, and 0.1 to 0.2 for the others. Adjust these values as necessary. Keep in mind that using elements effectively involves a process of trial and error.

Common elements combination

Here are some combinations of stylistic elements that have worked well for me when creating images of specific styles.

Cyberpunk/Futuristic combinations

Toon & anime + sparklecore + colourful Scribbles + solar punk
Dark Arts + Glowwave

Digital Painting combinations

Glowwave
Toon & anime + sparklecore + Digital Painting + colorful scribbles
Colorpop + Glowwave +sparklecore

HD stylized and cartoon style

CGI Noir + Colorpop + Oldschool Comic
Toon & Anime + Folk Art Illustration + Kids illustration +sparklecore

Fantasy and magical art

Colorpop + Glowwave +sparklecore
Glowwave + sparklecore

Easiest ways to enhance Leonardo AI images

Leonardo.ai offers Alchemy and Photoreal features which are a straightforward way to improve the quality of generated images. With Alchemy and Photoreal enabled Leonardo can add intricate details to images. Please note that both Alchemy and Photoreal are premium features.

Enhancing images with Alchemy

Alchemy is a powerful tool designed to enhance your images. It can increase the resolution, significantly improve the level of detail, and refine the lighting and shading in your images.

If you have a low-resolution image intended for printing or online sharing, enabling Alchemy can transform it into a cleaner, more detailed version. It’s also useful for enhancing existing images that lack sharpness or detail. Alchemy makes your pictures look even better, just how you want them. Plus, with Alchemy presets you can easily give your picture a certain look like Cinematic, Creative, Vibrant, and more.

Achieving Hyper-realism with Photoreal v2

If you want hyper-realistic and life-like images, enabling Photoreal v2 is the way to go. This feature can also be combined with the style transfer tool to produce consistently photorealistic photos.

Conclusion

To create better images in Leonardo.ai, it’s important to have a clear vision of what you want to achieve and to make the most of all the tools that Leonardo offers. Seek inspiration from other creators and don’t be afraid to experiment with the various settings available in Leonardo. By understanding how to effectively use the features described in this guide and combining them with your creativity, you’ll be able to produce stunning images with Leonardo AI.