How to make a 60-second AI explainer video, step by step

What makes a good explainer video

An explainer video has one job: answer a question the viewer didn’t know they had. The best topics are everyday things with a surprising answer. Why does a trackpad click when there’s no button? How do strangers help find your lost AirPods? People use these things every day and still can’t explain them. That gap is your hook.

Keep it to one question. If you have two, make two videos. That’s also why we aim for about 60 seconds, even though the platforms allow more. YouTube Shorts can run up to three minutes.¹ Instagram lets you record and edit reels up to 20 minutes, though it won’t recommend reels longer than three minutes to new audiences.² Sixty seconds is a choice. It forces you to cut everything that isn’t the answer.

The four-part structure of an animated explainer video

Every animated explainer video we make follows the same four beats. Here’s what each one does, with examples from our trackpad reel:

The hook is where most explainers go wrong. “Today we look at trackpads” is a topic, not a hook. A hook states something true that sounds impossible. Check it like any other fact: if the claim isn’t true, the whole video falls apart. We go deeper on openings in hooks for reels.

The payoff matters just as much. Flipping the opening hands viewers a sentence they can repeat. A closing line that names a specific person (“someone with a MacBook”) gives them someone to send it to. Skip the requests for likes and follows.

Write the script for the ear

An explainer video is heard before it’s read. Write it the way you’d explain it to a friend across the table:

• One idea per sentence, ideally 14 words or fewer.
• Concrete nouns and active verbs: “Sensors measure the pressure,” not “Pressure measurement takes place.”
• Repeat the noun instead of “it” when the picture changes. The viewer may have glanced away.
• Numbers as words in the voice-over, digits on screen as a label.
• Our rule of thumb: roughly 150 to 170 words for 60 seconds. Too long? Cut words. Don’t speed up the voice until it sounds rushed.

Then read it out loud with a timer. If you run out of breath, your viewer will too.

Every image on its word

This is the rule that makes an explainer feel sharp: when the voice says “sensor,” the sensor appears. Not half a second later. If the picture trails the word, the viewer is always catching up.

To place images that precisely, you need to know when each word is spoken. Text-to-speech services can tell you. ElevenLabs, for example, can return the audio together with a start and end time for every character of the script.⁴ Group the characters into words and you have a timestamp for every word. Now each image can start on its word, to the frame.

Two more habits keep the eyes busy: show something new at least every 1.5 seconds, and keep the camera moving. In a 60-second reel, a static frame feels long.

Pick a visual style and stick to it

Choose one look per series and write it down: background, colors, fonts, labels, how the camera moves. Our tech reels use clean 3D scenes. The trackpad reel is dark. The AirPods reel opens on a sunny café photo, which is AI-generated, then moves into a pastel 3D city. The look changes with the topic; the rules stay the same.

Be honest about what’s illustrated. If something is drawn bigger or simpler than it really is, a small “Simplified illustration” label costs you nothing.

Two ways to make an AI explainer video

You can split explainer video AI tools into two camps: some generate the pictures, others build them from code.

Route 1: a text-to-video generator

Here the AI explainer video generator does the picture-making: you describe a shot, and a model creates the footage. Google’s Veo 3.1, for example, makes realistic clips with natively generated sound and accepts reference images to keep a character or object consistent across scenes.⁵ That’s the fast route to realistic shots and quick drafts. For an explainer, check every word and number on screen before you post, because text in generated footage is part of the picture.

Route 2: code with Claude Code

Claude Code is an agentic coding tool: it reads your project, edits files and runs commands.⁶ For video, that means Claude writes each scene as code, and a renderer turns it into an MP4. HyperFrames, for example, renders HTML, CSS and animations into MP4 files and works with Three.js for 3D.⁷ On a Team or Enterprise plan, Claude Motion offers a code-based route inside Claude itself, in beta, with MP4 export.⁸

The code route takes more setup than typing a prompt. In return, every word, number and timing is exactly what you decided, and a fix touches only what you asked to change.

You can also mix the two: generate an image or a short clip, then place it inside a scene Claude builds. That’s how our AirPods reel got its café. For a deeper comparison, read AI video generator vs. Claude Code. If you want the code route ready to go, our course shows you how to make 3D explainer reels with Claude Code.

The pre-publish checklist for explainer videos

• Facts: every claim has a source you’ve actually read. Round honestly: “about,” “less than.”
• Hook: true, specific, about an everyday thing, and over within the first few seconds.
• Sync: every key word has its image, on its word. No stretch longer than 1.5 seconds without something new.
• Safe zone: keep text and key objects away from the edges. Meta’s ad specs for Instagram Reels recommend leaving about 14% at the top, 35% at the bottom and 6% on each side free of text and logos, so the profile icon and call-to-action don’t cover them.⁹ We use the same zone for every reel.
• Sound: the voice at a steady loudness, no clipping. Add music in the app.
• Labels: mark simplified illustrations. If you use realistic AI images or footage, follow each platform’s rules. YouTube asks creators to disclose realistic AI-generated or altered content, while clearly animated content and AI help with scripts don’t need it.¹⁰ Meta adds an “AI info” label when it detects industry signals of AI or when you disclose it yourself.¹¹
• Loop: the last frame matches the first, so the replay doesn’t jump.
• Format: vertical 1080 × 1920, watched once on a phone, with sound, before you post.

In our kit, Claude runs the mechanical checks itself before it hands over an MP4: loudness, the safe zone and the 1.5-second rule. That’s part of our Claude course for 60-second explainer reels.