Prompts are written instructions to help guide an AI model on what asset to generate. However, just like learning a new language, there’s a specific way to talk to the AI to get the results you’re looking for. You might have a specific vision in your head, but Layer can’t read your mind!

For example, many users might try the following for their first few prompts:

This is the basic building block for a prompt. But if you try using a simple prompt like this, you’ll notice that the output is highly varied, and will probably not be exactly what you’re looking for.

Good prompts don’t read like whole sentences. An OK prompt might read like this:

This actually isn’t so bad! You can see that additional descriptions can make things more accurate. However, let’s say we really want to pump up the visual fidelity, we would write something like this:

You can see that the generated results now have more focused visuals and increased quality of output. Additional “magic words” act like sweeteners and give more guidance to the model on what kind of image you’re looking for.

The words we added to our prompt here might seem like a laundry list of random words, but each one relates to a higher level of visual fidelity, even if they seem unrelated to the original prompt. For example, “highly detailed” is pretty self-explanatory, but “Houdini Render” is used because renders using the 3D art program Houdini tend to be very high fidelity (it’s used often for movies and AAA video games!).
​
You should adjust added words as needed to your prompts, but below are some example lists. Add the ones that feel right!

cinematic lighting, accent lighting, ambient lighting, backlight,

blacklight, blinding light, candlelight, concert lighting, crepuscular rays, direct sunlight, dusk, Edison bulb, electric arc, fire, fluorescent, glowing, glowing radioactively, glow-stick, lava glow, moonlight, natural lighting, neon lamp, nightclub lighting, nuclear waste glow, quantum dot display, spotlight, strobe, sunlight, ultraviolet, dramatic lighting, dark lighting, soft lighting, gloomy, foggy

highly detailed, grainy, realistic, unreal engine, octane render, bokeh, vray, houdini render, quixel megascans, depth of field (or dof), arnold render, 8k uhd, raytracing, cgi, lumen reflections, cgsociety, ultra realistic, volumetric fog, overglaze, analog photo, polaroid, 100mm, film photography, dslr, cinema4d, studio quality

Digital art, digital painting, color page, featured on pixiv (for anime/manga), trending on artstation, precise line-art, tarot card, character design, concept art, symmetry, golden ratio, evocative, award winning, shiny, smooth, surreal, divine, celestial, elegant, oil painting, soft, watercolor, washed color, fascinating, fine art

ultra wide-angle, wide-angle, aerial view, GoPro, massive scale, street level view, landscape, panoramic, bokeh, fisheye, long exposure, dutch angle, low angle, extreme long-shot, medium shot, long shot, close-up, extreme close-up, cowboy shot, medium cowboy shot, full shot, medium full shot, establishing shot, upper body, full body, highly detailed, depth of field (or dof), 4k, 8k uhd, photorealistic, ultra realistic, studio quality, octane render

Surrealism, trending on artstation, matte, elegant, illustration, digital paint, epic composition, beautiful, the most beautiful image ever seen

If your prompt starts getting long, you may see very varied results, you can emphasize or de-emphasize certain words using (parentheses) and +/- signs when forging.

If your generations frequently feature things you don’t like, try including them in [brackets] to refine your asset generation.

​