Sleepy Motion is an online motion graphics tool that turns a simple request into a finished, high-end video: an ad, a story, a news recap, a promo or a pitch. Not a slideshow, not a template with swapped text, but real kinetic typography, animated shapes, transitions, voice, music and sound effects synced to the millisecond. Usually ready in under two minutes.

You don't have to configure any of it. With AI decides, Sleepy reads your prompt and makes every creative decision for you: the mode, the look, the colors, the motion, the effects, even the voice. You describe what you want; it handles the how. And when you bring your own audio, photos or scene, nothing has to be generated at all.

At the top of the form you'll see two pills: AI decides and Custom settings. This is the most important choice, and for most people the answer is simple.

Start with AI decides, see a result, then switch to Custom settings to fine-tune or the other way around. Nothing is locked in.

Sleepy is a writer first. Whatever you ask for, it writes an actual script, so what you say shapes the entire narrative, tone, and angle of the video.

Be specific about the niche:

The more context you give location, clientele, what makes you different, the emotion you want to create the more the engine has to work with. Include brand personality words: bold, minimal, playful, premium, raw, clinical, vibrant.

Two things the Brain does that people miss:

• Just drop your website. “Make an ad for yoursite.com” the engine looks it up and builds the video around what you actually do. No need to write the pitch yourself.
• It knows what's happening. Ask for “a news video on the biggest story in tech this week” or “what happened today in the NBA” and it searches, finds it, and makes the video. Live and current.
• It speaks your language. Write your whole request in your own language, or just ask for one (“make it in Japanese”), and the video and voiceover come out in it.

Prompts that perform well:

• “A cinematic promo for a rooftop restaurant. Golden hour vibes, luxury dining, the city below.”
• “Hype video for a gym launch. Aggressive energy, fast cuts, motivational language.”
• “Soft product reveal for a new skincare line. Clean, minimal, aspirational.”

Just above the prompt box there's a small Want an idea? link. Open it for a menu of starter prompts grouped by what you're making trending & live, business & promo, local business, creator & social, teach & explain, motivation, and announcements.

Pick one and it drops a ready-to-run prompt into the box. The bits in [brackets] are yours to swap replace [product] or [topic] with your specifics and hit Generate. It's the fastest way to see what the engine can do, and a solid template for writing your own.

Inside the prompt box, bottom-left, there's a + button. That's where you hand the engine your own material the part only you can provide. Everything else it chooses.

• Use a preset: a ready-made look, instead of setting it yourself.
• Clone your voice: upload a sample or record one, and the video narrates in your voice.
• Upload voiceover: your own pre-recorded narration.
• Background music: your own track instead of one picked for you.
• Sound effects: up to 6 of your own clips for entrances and transitions.
• Text textures: custom image fills for your text.
• Upload logo: animated into the video, not slapped in a corner.
• Upload fonts: your own typefaces (TTF/OTF).
• Split-screen video: a clip to run under your content.
• Photos & video backdrop: images or a clip to run behind your content.

Cloning your voice? Once it's yours, drop cues like [laughs], [sighs] or [whispers] into your prompt and your voice will actually do them.

• Voiceovers in 22 languages. Write your request in your language or ask for one, and the narration follows: Arabic, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish and Turkish.
• Your own voice. Clone it from the + menu and it speaks every one of those languages, laughs and whispers included.
• Sound effects, on by default. Sleepy picks them and places them on every entrance and transition, synced to the motion to the millisecond, and varies them so they never sound repeated. Upload your own from the + menu if you have a sound you love.
• Music. Picked to match the mood and energy, or bring your own track.

Everything from here down lives under Custom settings optional, for when you want to steer. Skip it all and the AI Brain picks for you.

Brand colors are the highest-impact thing you can set. They're injected throughout the animation engine as the primary design system for that video used to set:

• Background color layers and gradient blends
• Text colors and contrast ratios
• Animated shape fills and strokes
• Overlay tints and transitions between screens
• Accent highlights on key words

When you don't set colors, the engine falls back to a neutral dark palette. That works but it will never feel yours. A video with your actual brand colors looks like something you paid for. One without looks like a demo.

Or set nothing and let Sleepy choose. If your prompt references a recognizable brand or a clear visual identity (“Apple-level minimalism”, “bold street brand”) it picks appropriate colors on its own.

No brand yet? The brand identity tutorial designs your colors, font, texture and text border with you in one AI chat.

Typography is half the visual identity of a motion graphics piece. Sleepy Motion ships with a curated font library but you can upload your own, and the engine will use it across all text layers in the generated video.

Uploading a font:

1. Hit the + in the prompt box and choose Upload fonts (or open the Fonts section in Custom settings).
2. Upload a .ttf or .otf file, any font you own a license for. Up to two. Font uploads are available from the Starter plan up.
3. Assign them as your Main Font and Accent Font.
4. The engine uses them across the text layers of that video.

Where to get fonts:

• Google Fonts free, massive selection, download the TTF directly.
• DaFont huge variety including display and decorative typefaces.
• Your brand kit if your business has a brand guide, it specifies the typeface. Download and upload that exact file.

When you upload a logo, the engine doesn't just slap it in a corner. It's treated as a motion element. Your logo enters with an animated transition a fade, a scale-in, or a directional entrance and is positioned intentionally within the composition.

The logo appears in multiple places across the video:

• As an intro or opening card element
• Anchored in a branded lower-third during narration screens
• On the outro / call-to-action screen

Logo format guidelines:

• PNG with transparent background this is the ideal format. Avoids white boxes over colored backgrounds.
• Square or nearly-square logos work best for most layouts.
• High resolution (at least 400×400px) the engine scales to fit but starts from the source.
• Add it from the + menu (Upload logo) or the Custom settings logo section.

Motion modes set the visual language of the whole video. The same prompt in two modes gives two completely different productions:

• Clean: crisp flat 2D graphics, the fastest (about 40 seconds).
• 3D and 3D Scene: the same clarity in 3D, or camera-driven 3D environments.
• Designer and Composite: poster cutouts, or images in portals with graphics inside.
• Hybrid, Liquid, Particles and Emoji: blended layers, fluid, morphing particles, emoji cutouts.
• Chat: fully customizable chat mockups.
• Sleepy Pro: motion graphics built into a scene video. See the Sleepy Pro tutorial.

Presets are one-tap looks (Liquid Fire, Glitch Paper, Pastel Glow and more), from the + menu. On top of any of it, post-process effects add a finish: deformers like liquid glass, halftone and print looks, color grades and film looks, up to 3 at once.

Vibe is a separate control that mainly shapes the voice delivery and the tone of the script: Calm, Cinematic, Corporate or Hype.

The vibe goes to both the scriptwriter and the voice, so delivery and written tone always match.

The two duration modes are structurally different, not just longer/shorter versions of the same output.

If you're unsure which to use: start at 10–15s. If the result feels rushed or cuts off before the message lands, switch to 25–30s.

Telling a longer story? Make several shorts, each in a different scene or mode, and put them together. Every part stays high-end, and it reads as one story.

• →

  Let AI decide first, then refine.

  Run one prompt on AI decides before touching a setting. See what the Brain gives you it's often already there. Then open Custom settings only for what you'd actually change.

• →

  Don't describe features describe the feeling.

  "Our app tracks calories" → generic. "The app people use when they finally decide to get serious" → a hook that actually works.

• →

  Use the Retry button with script edits.

  After your first generation, edit the script in the editor and hit Retry. You keep the creative direction and fix the specific lines that didn't land. Much faster than re-generating from scratch.

• →

  Match voice to vibe manually when needed.

  The Auto voice setting is tuned by vibe but if you know you want a specific voice for a specific brand, lock it in. Consistency across a campaign matters.

• →

  Short sentences in the description = short sentences in the script.

  The writer mirrors punctuation and sentence rhythm from your input. Punchy input → punchy script. Long run-on descriptions → meandering copy.

• →

  Brand colors are your highest-impact Custom setting.

  When you do go Custom, two brand colors turn a generic output into something that looks commissioned. Primary + accent is the sweet spot.

• →

  Split Screen (Brainrot mode) is not just for TikTok.

  It keeps viewers watching longer on any short-form platform by adding a second visual layer underneath. Instagram Reels and YouTube Shorts benefit just as much.

Live now:

• Sleepy Pro. Full motion graphics built into a scene video, from a preset, an AI video tool, your phone or Unreal Engine. Read the tutorial.
• Custom text borders. Design your own edge with Claude or ChatGPT and paste it in, the start of a full brand identity. Read the tutorial.
• 2K rendering. Toggle the small 2K badge next to the duration and your video renders at 2K instead of 1080p: sharper text, and it survives re-encoding when you upload it everywhere. Available from the Creator plan up (on Captions it just follows your source video).
• Voice cloning with expression, in 22 languages. Your own voice, and it can [laugh] and [whisper] on cue. See the + menu.