Layer is a creative AI platform for game studios and creative teams. Instead of wrestling with tools and settings, you describe what you want in plain language and Layer's assistant generates it — images, video, 3D, and audio — then helps you refine and organize everything in one workspace.

Think of Layer less like a single "generate" button and more like a creative collaborator: you have a conversation, it does the work, and your results collect on a visual canvas you can build on.

From the same chat, Layer can produce:

The heart of Layer is the chat. You tell the assistant — the Agent — what you want ("a glowing potion icon," "make three variations," "turn this into a video") and it works out the how: it picks a suitable model, sets the parameters, shows you the cost, and generates. You stay in the conversation to steer and refine, and the Agent keeps the context of your session.

💡 You don't need to know which model to pick or how to write a perfect prompt — describe the outcome and the Agent handles the rest. You can always get more specific when you want finer control.

Every chat session has a Board: a visual canvas beside the conversation where your results appear. Instead of scrolling a feed, you can arrange, group, and compare generations, zoom around, and build a piece of work visually.

A Reference Set is a named group of images that guides Layer toward a specific look, character, or style. Point the Agent at a reference set and your generations stay on-model and consistent — the same character, palette, or art direction across many assets.

Everything you make is saved automatically, so nothing gets lost:

You've got the lay of the land — the fastest way to understand Layer is to make something.