Most motion graphics sit on top of a video. Sleepy Pro puts them inside it. You bring a scene, and Sleepy masks it automatically and builds a full motion graphics production into the world: text walking through a desert, weaving between mountain peaks and clouds, passing behind the things in the shot.

The scene was never the hard part. Plenty of people can render a nice shot or generate one with AI. The hard part is putting animated text into it: masking every frame, tracking the camera, placing and timing each line so it moves behind and between things. Done by hand, that takes days, even for someone with real compositing skills.

Sleepy Pro does that part for you. Get a scene, upload it, and you have a finished high-end video in about 2 minutes, with no skills needed and no production budget.

1. Pick a preset scene. Our library scenes are rendered in Unreal Engine for the best possible quality: walks, landscapes, jobs, holidays and more. Pick one and you are done.
2. Make one with an AI video tool. The everyday route, and the most flexible. The next section walks through it.
3. Film it yourself. A steady walk, a product on a table, a view from a window. Even modest footage can give striking results.
4. Render it in Unreal Engine. For exact control over what goes where, we provide a sample project you can open, swap the model, move the camera and render. No Unreal experience needed, and a step-by-step guide is coming.

Every scene is reusable. Sleepy fills it differently on every render, so one great scene can carry a whole series of videos.

AI video tools are great at making a few seconds of a beautiful place. Sleepy Pro turns those seconds into a finished production. Here is the whole workflow:

1. Copy a scene prompt. In Sleepy Pro, under Make one with AI, copy the Turntable prompt (for a product) or the Walk-through prompt (for a place).
2. Generate the scene. Paste it into any AI video tool, fill in the [brackets] with your product or place, and ask for a vertical 4:5 video if the tool offers it.
3. Upscale it to 2K. Run it through an upscaler such as free.upscaler.video and pick the strongest detail settings. Detail is what Sleepy builds on.
4. Make it long enough. Most AI tools stop around 10 seconds. In DaVinci Resolve or any free editor, loop it, or add a reversed copy after it, to reach about 20 seconds.
5. Export at the exact size. Set the project to 2048×2560 (4:5), or 3072×3840 if you want a 2K video, and export an MP4 up to 100MB and 50 seconds.

1. In the generator, choose the Sleepy Pro pill in the motion modes, next to AI decides.
2. Under Your Video, upload your scene. Sleepy checks the size straight away and tells you if anything needs fixing.
3. Write your prompt as usual: what the video is about, in any language.
4. Generate. Your video is ready in about 2 minutes.

Want to decide exactly which parts of the scene the text goes behind? You can upload your own mask video as well, for example one rendered from Unreal. Leave it out and Sleepy decides, which often gives surprising results: in one of our tests, the text went behind a fence but in front of the zombies on the other side of it.

Here are four of our own scenes next to their masks. The mask marks what stays in front of the text, frame by frame, at the same 4:5 size as the scene.

• Post-process effects. Stack up to 3 high-quality effects, from liquid glass and halftone to film looks and color grades, and apply each one to the background only, the subject only, or everything. Beat-reactive effects hit on the bass.
• Sound effects. Added automatically and synced to every entrance and transition to the millisecond, or upload your own.
• Your look. Your fonts, brand colours, text border and texture all carry over. If you have not designed them yet, start with the brand identity tutorial.

With the same scene, a different prompt, effect or look gives you a completely different video. There are far more combinations than anyone could ever try.

Sleepy is very good at masking and placing text, so almost any scene works. The things that matter most are simple:

• Sharp. High detail and no motion blur. This matters more than anything else.
• Stable. Objects that stay solid. AI video sometimes melts or morphs things, so re-generate any scene where that happens.
• Something to work with. Clear things in the shot, such as a product, rocks, plants, buildings or a fence. Several are fine.
• One continuous camera move. No cuts. A little open space on the sides helps, but it is not essential.

2D illustrated scenes work too, and look great with motion graphics.

• Every plan, including Free: a few preset scenes, in HD.
• Creator and up: every preset scene, uploading your own scenes, and 2K output.

Try a free preset scene first to see what Sleepy Pro does. When you want your own scenes, Creator unlocks the whole workflow.