Script Timing Math: How Many Words Fit in a 10s, 15s, 30s, and 60s Video?

1. Convert words per minute into a usable time estimate

Words per minute expresses a rate over sixty seconds. Divide the rate by sixty to get words per second, then multiply by the video duration to calculate a continuous-speech capacity. At 150 WPM, the rate is 2.5 words per second. Twenty-five words therefore take ten seconds in the model. Reversing the calculation gives estimated seconds as word count multiplied by sixty and divided by WPM. These two versions are useful for different decisions: one starts with an edit length, while the other starts with an existing script.

When the result includes part of a word, round the capacity downward if you are setting a hard whole-word limit. At 150 WPM, fifteen seconds allows 37.5 theoretical words, so the timer's cutoff is thirty-seven whole words. That does not mean word thirty-eight is always impossible in a recording. It means the tool uses a conservative, reproducible token boundary rather than displaying a fractional word as something a writer can deliver. Keep the exact decimal time estimate separate from the whole-word editing threshold.

2. Subtract silence, demonstration time, and the final hold

A video's duration is not necessarily its speech duration. If a fifteen-second tutorial needs a one-second opening demonstration, a half-second pause before the reveal, and a one-and-a-half-second final hold, speech has twelve seconds left. At 150 WPM, the capacity becomes thirty words rather than thirty-seven. The tool's main estimate does not automatically subtract those production choices, so make the allowance in your own plan. This is especially important for a visual explanation where the viewer must inspect a diagram or compare two states without hearing a new instruction immediately.

Do not distribute silence mechanically between every sentence. Reserve it where the audience needs to process a visual change or where the voice performance needs emphasis. A short pause before revealing a marked fill line has a different purpose from dead air before anything interesting happens. Write those holds as timeline events instead of burying them in punctuation. A comma in a text field does not specify a fixed number of milliseconds, and different text-to-speech systems may interpret the same punctuation differently.

const duration = 15;
const nonSpeechSeconds = 3;
const wpm = 150;
const capacity = Math.floor((duration - nonSpeechSeconds) * wpm / 60);
// 30 words, leaving 3 seconds for demonstration and holds
const estimate = (wordCount) => wordCount * 60 / wpm;
// estimate(30) === 12 seconds; recording still required

3. Understand what the counter measures and what it does not

The Script Timer counts English word tokens separated by whitespace. A contraction such as don't is one token, while a phrase containing spaces creates several. A written number such as 1080 is one token even if a narrator expands it into several spoken words. The same problem appears with URLs, initials, formulas, and product names. Before timing a final script, replace those elements with the words you actually intend to say. Count the performed sentence rather than relying on the shortest possible notation in the document.

The Prompt Builder uses a separate word-oriented estimate for its individual dialogue beats. Punctuation-heavy or unusual input can therefore produce a small difference between the two displays. Use the Script Timer for a consistent whole-script comparison and the audio track for the final edit. Neither counter is a pronunciation engine, a speech recognizer, or a multilingual timing model. The visible character count is also not a promise about perceived symbols in every writing system. For non-English scripts, measure a real read rather than applying the English WPM table uncritically.

4. Use the cutoff word as an editing signal

Choose a target duration and WPM, then inspect the point where the script exceeds the calculated word capacity. That cutoff is a location to investigate, not an instruction to delete the final phrase blindly. Often the most useful fix is earlier: remove an introductory clause, replace a long explanation with a visible demonstration, or merge two sentences that repeat the same idea. Protect the actual result and the instruction the viewer needs. A script that fits but omits the promised answer is not a successful revision.

For a thirty-second explanation at 150 WPM, start with a seventy-five-word mathematical ceiling and leave room for the visual plan. If the first draft is ninety words, do not simply choose 180 WPM because the number turns favorable. Read both versions aloud. The faster setting may work for a familiar list, but a tutorial with unfamiliar terms needs additional processing time. Use rate changes as deliberate performance choices, not as a way to avoid prioritizing the material. The viewer cannot pause your pacing decision while watching a continuous demonstration.

5. Allocate words according to what each beat must explain

A ten-second, three-beat story does not require an equal number of words in every section. The opening might need only three words because the contradiction is visible. The middle may need five words to name the consequence, while the ending needs seven to deliver the actionable distinction. At 150 WPM, those fifteen words occupy about six seconds in the estimate, leaving four seconds across the edit for visual reading and natural pauses. Let the action determine where language is valuable rather than filling each beat to its maximum capacity.

A sixty-second instructional video needs a different allocation. Reserve a concise opening promise, enough time for the actual steps, and a clear finished result. If you spend twenty seconds explaining why the topic matters, the remaining forty seconds may not support the procedure you promised. Map each sentence to a specific visible state or operation. Any sentence with no visual role should justify its place through essential context. This is a practical way to cut words without turning a useful explanation into an unexplained montage.

6. Validate with a scratch read before final production

Record a rough version using the actual intended delivery: neutral, reflective, energetic, or instructional. Measure from the first audible word to the final word, then note any pauses you want to preserve. Compare that duration with the calculator's estimate. If your thirty-word script takes fourteen seconds instead of twelve, the recording is not wrong. Your delivery simply differs from the 150 WPM model. Use the observed timing to revise the edit or the script rather than forcing the waveform into a planning assumption.

When using a synthetic voice, generate the actual chosen voice with the final punctuation and pronunciation settings. Changing the voice later can alter the timing even when the text stays identical. Save the audio version with the corresponding script and cue sheet. Mark the onset of key nouns and action verbs in the editor, then check whether the corresponding object or action is visible. This catches synchronization errors that a total duration cannot reveal. A twelve-second file can still describe the wrong thing at the wrong moment.

7. Keep timing changes auditable across revisions

Store the target duration, selected WPM, non-speech allowance, script word count, and actual recorded duration in a short production note. When feedback arrives, you can distinguish a content request from a pacing request. Adding one explanatory sentence might require removing another sentence or extending the sequence. Without a budget, every small addition seems harmless until the final delivery feels crowded. Version the script and audio together so an editor does not accidentally assemble a new storyboard against an obsolete read.

Finish with a phone-sized playback at normal speed. Check whether captions remain readable while the speaker continues and whether the last visual stays long enough to register. Do not mistake an empty final second for wasted time if it lets the audience recognize the result. Conversely, remove silence that serves no visual or performance purpose. The calculator's value is making the first constraint visible; your final responsibility is to balance spoken information, visual information, and the viewer's opportunity to understand both.

Frequently asked questions