Guide

Explainer video words per minute

An explainer video voiceover runs at about 130 to 150 words per minute. Plan at 150, which is 2.5 words a second, then let the recorded voice set the real timing. Here's how many words fit each length, how the pace decides where the cuts go, and how to tighten it with your AI agent and FrameJam.

How many words per minute for an explainer video

Script guides like VideoScribe's put a comfortable voiceover at 130 to 150 words a minute. FrameJam's explainer playbook plans scripts at 2.5 words a second, which is 150 words a minute, before any audio exists. Treat the numbers below as a ceiling, not a quota:

Video length	Calm, 130 WPM	Default, 150 WPM
30 seconds	65 words	75 words
60 seconds	130 words	150 words
90 seconds	195 words	225 words
2 minutes	260 words	300 words
3 minutes	390 words	450 words

The playbook's default explainer is about 90 seconds at 16:9, so about 225 words. For X it stays at 2:20 or less, and anything between 1 and 3 minutes works.

What makes it faster or slower

New ideas need room. A technical topic, a new term or a number the viewer has to remember reads better nearer 130 words a minute.
Diagrams need holds. If a scene builds a chart or a mechanism, leave a beat at the end where nobody talks and take those seconds out of the word budget.
Numbers, names and links take longer to say than they look. Write them the way they're spoken ("two point five", not "2.5") and count them that way.
Vertical shorts run brisker, but cut words before you speed up the voice. A clear voice at 150 beats a rushed one at 180.

How the pace decides the cuts

In a narrated explainer the voice is the clock. Every cut, drawing and caption hangs off the word that names it, so the script's pace becomes the edit's pace. The explainer playbook works with these rules:

One idea per scene, 5 to 15 seconds long, with 1 to 4 short sentences and 12 to 45 words.
Write in cues: short phrases with clear breaks ("First the snowball. Then the hill. Then the speed."). Each phrase becomes a reveal; a long run-on sentence gives the picture nothing to land on.
Every picture appears within about a fifth of a second of its word, never before it.
Each scene ends on a short held read before the next one starts, so the viewer can take it in.

Explainer text on screen

The voice carries the sentences; the screen carries the key words. Keep each on-screen element to 7 words or fewer, captions at least 40 px tall so they read on a phone, and nothing under the caption band. If a scene needs a paragraph on screen, it's two scenes.

Explainer illustrations that land on the word

Faster narration means more pictures per minute, so invent one clear visual idea per scene: a metaphor drawn as an object, a diagram that grows, a number with its takeaway, or an original mascot that holds each step's prop. The look comes from a style, like the technical lines of Blueprint, a friendly Doodle Mascot or hand-drawn Paper Marker. More looks, each with a prompt, are in the free Remotion templates gallery.

Tighten the timing with your agent and FrameJam

You don't have to time anything by hand. With FrameJam installed in Claude, ChatGPT or Cursor, your agent follows the explainer playbook:

It writes the script against the length you asked for, at about 2.5 words a second, split into scenes with a cue on every word where something should appear.
It records or generates the voice first (your own recording, or Kokoro, a free voice that runs on your computer) and times every word with whisper.cpp. Those word times become the clock for the whole video.
It animates each scene on its cues, renders, and opens the video in FrameJam. Want to see the scenes first? Ask for a storyboard.
You review it on exact frames: drag across a scene that drags and write "too slow, cut this to two sentences", or pin a drawing that arrives late. Your agent maps each comment to its scene and cue, re-voices only that scene, and everything placed on its cues moves with it.

Use the FrameJam explainer playbook to make a 90-second explainer on how compound interest works, for people new to investing. Keep the narration at about 150 words a minute and show me a storyboard first.

Already have a script? Ask for a timing check before anything is animated:

Here's my explainer script: ~/Documents/script.md. How long does it run at 150 words a minute? Trim it to 60 seconds without losing the main idea, then make the video with the FrameJam explainer playbook.

Prefer a picker? Type /framejam:playbook in your agent to choose a playbook on a page, or build the prompt with the AI video prompt generator. The same voice-first timing drives the 3Blue1Brown math animation playbook, and the how-to video playbook uses it when you add narration.

Vertical explainers for Shorts and Reels

A 9:16 explainer has less room for text and loses the bottom of the frame to the app's captions and buttons. Keep scenes short, put the hook in big words in the first three seconds, and check your frames with the YouTube Shorts safe zone checker before you post.

Questions

How many words per minute is an explainer video?

About 130 to 150 words a minute for the narration. 150 words a minute, or 2.5 words a second, is a good default: a 60-second explainer holds about 150 words and a 90-second one about 225.

How many words is a 2-minute explainer video script?

About 260 to 300 words at 130 to 150 words a minute. Take some off if the video needs silent holds for diagrams or a product on screen.

Should the voiceover be faster for Shorts and Reels?

A little. Vertical explainers usually run brisker, and the hook has to work in the first three seconds with the sound off, so lead with big words on screen. Don't speed up a voice that's already clear; cut words instead.

How much text should be on screen?

Very little. FrameJam's explainer playbook keeps each on-screen element to 7 words or fewer and captions at least 40 px tall, so the viewer listens to the voice instead of reading paragraphs.

Does FrameJam write the script?

Your agent does, following the explainer playbook, or it keeps your own script word for word. FrameJam is where you review each render and point at the exact moments to change.

Works with

To build any video from scratch with your agent, read how to make videos with Claude; to cut a recording instead, read how to edit videos with Claude. Explainers in Hyperframes start from the Hyperframes Claude skill.