Yes, one sentence is enough to start an animated video with AI: the tool writes the narration, splits it into scenes, illustrates it, and adds the voice and captions. But the sentence decides everything. "Cats" gives a vague video; "Why cats purr, and not only when they're happy" gives a video that holds up. The structure that works best fits in three lines: the topic as a question, two or three checked facts, an angle. It's the format of the topic examples offered in ClipOnIt's Explainer.
What happens between your sentence and the video
When you give the Explainer an idea, several steps run without you:
- Writing the narration: a few short sentences, answer first.
- Splitting into scenes: one sentence per scene, 3 to 8 seconds each.
- Composing the visuals: for each scene, AI picks and places animated elements in the chosen style, timed to the words of the voice.
- Voice and captions: one of the catalog voices reads the text; captions follow the voice.
- Render: a vertical 9:16 MP4, in about 4 minutes.
At every step, AI fills in what your sentence leaves out. The vaguer the sentence, the more it fills with generic material.
Example: "Why cats purr"
This 19-second video was rendered by the Explainer in the Mood pixel style (warm, glowing pixel art). Its script is 4 sentences and 61 words; it was reread and approved before rendering, as every script should be. Voice-over: Mathis, ClipOnIt's founder.
Your cat purrs on your lap. And yet, it's not just a sign of happiness. It starts in the throat: muscles vibrate twenty-five to a hundred and fifty times per second. Kittens purr from just a few days old to talk to their mother. As adults, they also purr when they're stressed or hurt. These vibrations might even help heal bones.
Look at what this text contains: a familiar situation (the cat on your lap), a surprise (more than happiness), a figure (25 to 150 vibrations per second), an explanation, and an ending that opens up ("might even…"). All of that fits in a three-line request.
The three-line structure
The Explainer's topic examples all follow the same format. Here it is, applied to cats:
Topic: why cats purr.
Key facts: throat muscles vibrate 25 to 150 times per second; kittens purr from a few days old; adults also purr when stressed or hurt.
Angle: it's more than a sign of happiness.
The topic, as a question
Write it the way someone would type it into a search bar: "why…", "how…", "how much…". A question calls for an answer, and the answer becomes the first sentence of the video.
Two or three facts you have checked
This is the most important line. Without facts supplied, AI writes the ones it thinks it knows; they're often right, sometimes wrong, and you won't know which. By supplying them, you control what the video claims. One precise figure per video is enough.
An angle
The angle says what makes the video worth watching: a misconception to fix, a concrete consequence ("what it means for your morning cup"), a surprise. Without an angle, the video recites a definition.
Sentences that work, sentences that miss
| Starting sentence | What you get | Better |
|---|---|---|
| Cats | A list of generalities | Why cats purr, even when stressed |
| Inflation | A textbook definition | Why your rent rises faster than your salary |
| Our new app | An announcement with nothing to hold on to | The problem our app solves, in one sentence, for whom, at what price |
| QR codes | A history lesson | Why a damaged QR code still works |
The "Product launch" topic example in the Explainer follows this logic: what the product does, who it's for, the price, the date, and the problem it solves in one sentence.
After generation: three ways to fix things
- The AI Director: type or dictate "shorter", "add a figure here", "change the ending". Each proposed change is shown before it's applied; you accept or decline it, and everything can be undone.
- Variants: three other versions of a scene, side by side, with the same text read.
- By hand: move, hide or lock an element, rewrite a sentence.
If you start from a finished text rather than a sentence, the method differs: see turning a text into an animated video.
And from Claude or ChatGPT?
The Explainer is also available through an MCP server and an API from the Starter plan: you can ask Claude or ChatGPT "make an explainer video on why cats purr, mood pixel style", and the assistant calls the production tool end to end. The rule stays the same: give it the facts and the angle. How MCP works is covered in automating your shorts with ChatGPT or Claude.
What a sentence doesn't replace
- Checking: reread the script before rendering, especially figures and dates.
- Choosing the topic: a well-made video on a question nobody asks won't be watched.
- Consistency: same style, same voice from one video to the next is what makes a channel recognizable.
What it costs
The Explainer is included from Starter: $29/month for 15 explainer videos a month, $49/month for 30 (Pro), $149/month for 100 (Business). The Free plan does not include explainer renders. See the Explainer page and its 22 visual styles, plus your own.
FAQ
Can you really make a video from a single sentence?
Yes, the Explainer accepts a one-line idea and produces a complete video. Results are better with three lines (topic, facts, angle), and you should always reread the generated script.
How long is a video made from one sentence?
Usually 20 to 60 seconds, or 4 to 10 scenes. Our English examples run from 19 to 45 seconds. If you want it shorter, say so in the request or ask the Director afterwards.
Does AI make up facts?
It can be wrong, like any language model. That's why the "key facts" line exists: the video relies on what you have checked.
Do I pick the style first?
Yes: the style is chosen at the start, because it decides which elements AI looks for in the library. To choose, see which explainer video style for which topic.
Written by the ClipOnIt team. Published November 7, 2026.