An educational video made with AI is most useful at three moments: opening a lesson with a question, reviewing a concept in one minute, and preparing a flipped classroom. The method that works: one learning objective per video, 30 seconds to 2 minutes, one picture per idea, your own voice if possible, and a recall question at the end. AI does the production (scenes, animated illustrations, voice, captions); the content, the accuracy of the facts and the order of concepts stay your job. ClipOnIt's Explainer produces this kind of video in vertical 9:16, from $29/month for 15 videos.
What a short educational video is for
A short video doesn't replace a lesson. It's a way in, or a reminder. Four uses come up again and again among teachers and trainers:
- The lesson hook: a surprising question in 30 seconds ("why do we yawn when someone else yawns?") to open a unit on the nervous system or mirror neurons.
- Review: one concept, one minute, to rewatch on a phone before a test or a certification.
- Flipped classroom: learners watch the explanation before class; class time goes to exercises and questions.
- Corporate micro-learning: a procedure, a safety rule, a software feature, inside a training module.
Example: "Why yawning is contagious"
This 29-second video was produced by the Explainer in the Class sketch style: hand-drawn, black and white with red accents, like a whiteboard. Voice-over: Mathis, ClipOnIt's founder.
You'll yawn before this video ends. Yawning is a reflex. Scientists still don't know exactly what it's for. One idea: it may help cool the brain. Another: it may help you wake up a little. But the strangest part is the contagion. […] So, did you yawn?
Three teaching choices visible in this script:
- It says what we don't know. "Scientists still don't know exactly" is honest, and it shows that science moves through hypotheses ("one idea… another…").
- It involves the learner from the first sentence ("you'll yawn") to the last ("did you yawn?").
- It qualifies its sources: "observations suggest" rather than "it is proven that".
The method in 6 steps
1. One learning objective per video
Phrase it the way you would in a lesson plan: "by the end, the learner can explain why…". Two objectives means two videos.
2. Start from the question, not the syllabus
Turn the objective into a question a learner might ask. "Mirror neurons" becomes "why do we yawn when someone else yawns?". The answer comes in the first seconds, the explanation after.
3. Write short, for the ear
Count on about 170 to 190 words per minute in English. Sentences of 8 to 15 words, one technical term at a time, defined when it first appears. If you start from lesson notes, the rewriting method is in turning a text into an animated video.
4. One picture per idea
Each scene shows what the voice says, when it says it. Illustrations are not decoration: a diagram that appears on the right word helps memory, an off-topic picture makes learners lose the thread.
5. Your own voice, if you can
Your learners know your voice. In the Explainer, you can record the narration on the mic (one or several takes) or import an audio file: it's transcribed word by word, and each illustration is timed to your words. A fluffed sentence can be re-recorded on its own. Otherwise, 4 English voices and 6 French voices are available.
6. End with a recall question
A question the learner answers in their head ("so, did you yawn?", "what's the difference between… ?") does more for memory than a summary. You can pick it up again at the start of the next lesson.
Which style for which subject?
| Subject or use | Explainer style | Why |
|---|---|---|
| Science, biology | Class sketch | Whiteboard look, readable diagrams |
| Geography, history, economics | Satellite briefing | Real satellite imagery, routes, map pins |
| Physics, technology | Cutaway house, Wireframe | Cutaway objects or wireframe 3D |
| Corporate training | Office toon, Flat vector | Office characters, clean illustration |
| Key figures, statistics | Swiss grid | Giant figure, strict grid |
These are suggestions, not rules. The full list is in which explainer video style to pick.
What to know before you start
- Vertical only. The Explainer renders in 9:16 (1080×1920), made for phones, a learning platform viewed on mobile, or social media. For 16:9 classroom projection, it isn't the right tool today.
- Check every fact. AI can get a date or a figure wrong; an error in a lesson video spreads to every learner.
- No personal data about students in texts or recordings. ClipOnIt videos are hosted in the European Union (Hetzner, Finland); processors are listed in the privacy policy.
- Usage rights: the videos you render belong to you and can go wherever you want (Moodle, an LMS, YouTube, social media).
Time and cost
For a 30 to 60-second video, expect 10 to 30 minutes of work: most of it goes to writing and reviewing; rendering takes about 4 minutes. These are orders of magnitude, not a measurement.
| ClipOnIt plan | Price | Explainer videos / month | Transcribed voice minutes / month |
|---|---|---|---|
| Starter | $29/month | 15 | 60 |
| Pro | $49/month | 30 | 180 |
| Business (3 seats) | $149/month | 100 | 600 |
Source: plan catalog as of November 7, 2026, see pricing. The Free plan does not include explainer videos. For a whole course, see also turning a course into explainer videos and the Explainer page.
FAQ
Can an AI educational video replace a lesson?
No, and it isn't meant to. It opens a lesson, reviews a concept or prepares a flipped classroom. In-depth explanation, exercises and discussion stay in class.
How long should an educational video be?
30 seconds to 2 minutes per concept. Beyond that, split it: a series of short videos is easier to review than one long one.
Do I need to show my face?
No. An animated illustration with your voice is enough, and it often shows the mechanism better than a face would. See faceless explainer videos.
Can I make the same video in French for a language class or an international audience?
Yes: duplicate the project, rewrite the text in French and pick one of the 6 French voices, or record yourself. Each version counts as one video in the quota.
Written by the ClipOnIt team. Published November 7, 2026.