Estimated duration equals spoken words ÷ words per minute × 60, plus silent pauses and any separate end-card hold. This is a planning estimate. Time the generated or recorded voiceover before locking the edit.
Try your script.
A worked 15-second brief.
The sample contains nine spoken words. At the illustrative setting of 120 words per minute, narration takes 4.5 seconds. Add two seconds of pauses and a two-second silent closing hold: the estimate is 8.5 seconds, leaving 6.5 seconds within a 15-second target.
That remaining time is creative space, not a requirement to add more words. You could show a product detail or let the viewer read the closing action. If the voice continues over the closing card, count that time as narration and do not add the same seconds again as a silent hold.
Measure the pace of your own voice.
Use a short recording in the intended voice, language and delivery. Count the words actually spoken, spelling out numbers and abbreviations as they were read. Time from the start of the first word to the end of the last word, including natural pauses within that passage. Exclude leading silence and a silent closing card.
Worked example: nine spoken words over six seconds give 90 words per minute. Applying that pace to the nine-word script above gives six seconds of narration. Two additional seconds of silence and a separate two-second closing hold bring the estimate to ten seconds. These are illustrative measurements, not a production benchmark.
The measured pace already includes natural pauses in the sample. Add only extra silence planned outside that narration; do not add those pauses again. A different script, voice or language may need another sample. The estimate is still a planning aid, and the finished recording must be timed directly.
Return to the script timer ↑ · Use the pronunciation brief to settle names and reading choices before timing.
Use a pace you can defend.
The default is an editable example, not a universal recommendation or a measured Aitoma production benchmark. Read the draft in the intended voice, or generate a short test. Divide the number of spoken words by its duration in minutes to estimate that particular delivery’s pace.
Brand names, numbers, abbreviations, emphasis, and language can change duration. Write numbers as they will be spoken before calculating. “24/7” may be one written token but several spoken words. This tool counts whitespace-separated words and is not suitable for languages whose words are not separated by spaces.
Shorten the message before speeding it up.
| If the estimate exceeds the target | Useful revision |
|---|---|
| Two competing promises | Choose the promise that answers this viewer’s question. |
| A long feature list | Keep the feature needed to explain the use case. |
| Repeated on-screen and spoken instructions | Let the visual carry a detail when comprehension remains clear. |
| A rushed final action | Reserve time to read and recognize the destination. |
For example, “Our innovative coffee collection gives you a range of different options to discover the perfect coffee for your morning routine” can become “Find the coffee for your morning.” The shorter wording changes the message and still needs to match the product.
Verify against the final audio.
Once the voiceover exists, place it in the timeline and measure its actual duration. Check the opening, pauses, caption changes, and ending at normal speed. Timing estimates cannot detect mispronunciation, unnatural delivery, or text that disappears too quickly.
Use a storyboard to assign visual actions to the measured audio. The final export, including music and transitions, is the version to approve. A script that fits the time budget can still be hard to follow if the edit introduces too many ideas.