How to Write a 60 Second Video Script (With AI)

Published June 27, 2026 · Updated June 27, 2026 · 6 min read · by the Sakura5 team

A 60-second video script is about 150 words. That's the whole secret most creators are missing: they write 300 words, start recording, and wonder why every take runs long and rushed. Nail the word count first and the rest of the process gets dramatically easier.

The math comes from speaking pace. Most people deliver conversational on-camera speech at roughly 140–160 words per minute — call it 150 words per minute of speech as a working rule. A 30-second video is 70–80 words. A 90-second video is 210–240. Write to the budget, not to the idea's natural sprawl.

Below is the structure that makes those 150 words work — hook, body, CTA with word budgets for each — plus three fill-in templates and how AI scripting collapses the whole process into a couple of minutes.

The Words-Per-Minute Math

Before structure, budgets. At ~150 words per minute:

• 15 seconds ≈ 35–40 words
• 30 seconds ≈ 70–80 words
• 60 seconds ≈ 140–160 words
• 90 seconds ≈ 210–240 words

Two adjustments. First, energy: punchy, hyped delivery runs faster (closer to 170 wpm), deliberate explainer delivery runs slower (down toward 120). Second, pauses: every beat you hold for effect costs words from the budget. If your style leans on pauses, write 10 percent short.

Time your own pace once — read any 150-word passage aloud at performance energy and check the clock — and you'll never guess again.

The Hook–Body–CTA Structure

Hook: 10–20 words, the first 3–5 seconds

The hook has one job: give a viewer a reason not to swipe. No greeting, no "in this video I'm going to." Open with the tension, the promise, or the surprising claim itself. If your first sentence could be deleted without losing information, delete it — the second sentence was your hook all along.

Body: 100–115 words, one idea in three beats

Sixty seconds fits exactly one idea. Not three tips and a story and context — one idea, developed in roughly three beats of 30–40 words each. Common beat patterns: problem → why it happens → the fix; claim → proof → how to apply it; step one → step two → step three. Each beat should be speakable in a single breath or two.

CTA: 15–20 words, one ask

End with exactly one action: follow, comment a keyword, save the video, tap the link. Two asks halve each other. Tie the ask to the value: "Follow for the next one" works because the body just proved the follow is worth it.

Three Fill-in Templates

Copy these, fill the brackets, and check the word count against your budget.

Template 1 — the mistake reversal:

Stop [common mistake] if you want [result]. Here's the problem: [why the mistake backfires, one sentence]. Do this instead: [step one]. Then [step two]. The difference is [payoff in one concrete sentence]. Follow for more [topic] like this.

Template 2 — the numbered payoff:

Three [tools/habits/settings] that [specific result] — and number two is the one everyone skips. One: [item plus a one-line reason]. Two: [item — spend an extra sentence here, it's the star]. Three: [item plus one line]. Start with number two today, and [expected outcome]. Save this so you don't lose it.

Template 3 — the story open:

[Time frame] ago, I [surprising or relatable event]. I thought [initial assumption] — I was wrong. What actually happened: [the turn, two sentences]. The lesson: [one-sentence takeaway you can act on]. If you've dealt with [same situation], comment [keyword] and I'll share the full breakdown.

How AI Scripting Speeds This Up

Everything above is mechanical — word budgets, beat structure, one ask — which is exactly why AI is good at it. The catch is that generic chatbots don't think in seconds; they think in paragraphs, so you end up trimming an essay down to 150 words by hand.

Purpose-built tools skip that step. In Cueflow, you give the AI an idea, pick a tone, and set an exact target length; it returns a script sized for that runtime. It also learns your style and your signature hook as you use it, so drafts open the way you open, not the way a press release does. While you edit, an estimated reading time updates live, so "does this fit in 60 seconds" is answered before the first take — and then the same app scrolls the script over your camera near the lens while it records.

The workflow becomes: idea → generated draft → 60 seconds of trimming in your own voice → record. And once you're recording, delivery technique takes over — here's how to read a script on camera without looking like you're reading.

FAQ

How many words is a 60 second video script?

About 140–160 words at a conversational on-camera pace of 140–160 words per minute. Budget roughly 10–20 words for the hook, 100–115 for the body, and 15–20 for the call to action, and write slightly short if your style uses pauses.

What makes a good hook for a short video?

A specific promise, tension, or surprising claim delivered in the first 10–20 words, with no greeting or preamble. If deleting your first sentence loses nothing, cut it — the strongest opening line is usually the one you wrote second.

Can AI write a video script that sounds like me?

It can if the tool is built for it. Look for AI scripting that lets you set tone and exact length and that learns your style over time — Cueflow, for example, picks up your signature hook from use, so generated drafts open the way you naturally do.

← All articles