Skip to main content

How do I create a Custom Style in Layer?

Learn how to train a custom AI style on FLUX, SDXL, or BRIA on Layer - letting you keep game assets consistent and high quality.

Updated over 2 weeks ago

Training a custom style (aka LoRA training) in Layer lets you generate assets that match a unique visual style - whether it’s a specific game art direction, a hand-drawn look, or something entirely custom.

This guide shows how to train your own custom style in Layer, step by step.


Step 1: Start a New Style

Once your training assets are ready (cropped, cleaned up, and consistently formatted), head over to the Art Styles page.

At the top center, click “Build a New Style.” This will kick off the guided style creation flow - creating a new style for your workspace.

If you’ve trained a model somewhere else and want to bring it into Layer, select “Bring in a model.”

Otherwise, hit Get Started to begin setting up a new style from scratch.


Step 2: Set Up Your Style Details

On this screen, you’ll define the core info for your style:

Style Name

Give your style a name that helps you recognize it later.

Style Base

Choose which foundational model you want your style to be trained on.

Style Type

This tells Layer what kind of assets your style is meant to generate. Pick the option that best matches your use case.

Base Styles Explained

  • Stable Diffusion XL (SDXL) – An open-source foundational AI model. Great for a wide range of use cases and highly customizable.

  • BRIA – A copyright-compliant, privacy-safe foundational AI model. Ideal for users with stricter legal or ethical guidelines, especially for commercial workflows.

  • FLUX – A high-performance model developed by Black Forest Labs, known for its exceptional visual quality, prompt adherence, and style diversity.

Style Types Explained

  • Single Character – For creating poses, angles, and varying shots of characters.

  • In-Game Items – For items and equipment like trophies, gems, weapons, etc.

  • Backgrounds – For game backgrounds, including paintings, isometrics, & more.

  • Multiple Characters – For variations of a character type, like avatars same styles.

  • Vehicles – For generating different vehicle types (cars, ships, bikes, etc.).

  • Environmental Objects – For props/structures like houses, trees, and scenery.

  • UI – For buttons, frames, menus, and other interface elements.

  • Icons and Symbols – For icons and visual symbols in a consistent style.

  • Other – If none of the above fit what you’re working on.


Step 3: Upload Your Training Assets

For guidance on the best assets to upload when creating a style, check out our video tutorial:

Now it’s time to upload your images. You have two options:

  • Upload them directly on this step

  • OR (if your images are already in your Layer Drive) select them all, right-click, and choose “Create style from files”

Once uploaded, Layer will auto-caption each image. These captions describe what’s in the image and help the AI learn what defines your style.

💡 Pro tip: Review the captions. They’re usually solid, but sometimes they miss key details or mislabel things — especially with complex images. You can edit the captions directly in the interface.


Step 4: Add Evaluation Prompts

Here you’ll add at least 3 evaluation prompts to show what kinds of assets you’ll be generating with your style.

These don’t affect the actual training, but help preview what the results might look like once the style is ready.

Think of these as early test prompts, such as:

  • “A cute forest house with a chimney”

  • “A knight holding a glowing sword”

  • “Top-down view of a treasure chest”

Keep them simple and aligned with your intended use.


Step 5: Train and Wait

Once everything is set, kick off the training.

Depending on the base model you selected, training can take as little as 15 minutes. You’ll get notified when it’s done.


You’re Ready to Forge

After training completes, your style is live. You can now use it in Forge to generate style-consistent assets, matched to the look and feel you trained the model on.

Did this answer your question?