Voiceover Words per Minute: Arabic vs English, with 15, 30 and 60-Second Word Counts

01Voiceover words per minute: the quick numbers

"How many words should I write?" is the first question behind almost every ad script. Most figures online come from English and get applied to Arabic unchanged, so the Arabic spot either overruns the slot or gets read so fast it suffocates. These are the planning figures Vocla's tools use, at a natural pace:

The last three columns include a short breath after each sentence, which is why they come in slightly under simple division. Note the gap between English and MSA: about 18 words in a 30-second spot. Keep the English word count for your Arabic version and it will run closer to 38 seconds than 30.

Figures are planning estimates from Vocla's voiceover length calculator as of October 2026: a words-per-second rate per dialect plus a pause of about 0.3 seconds between sentences. For comparison, the audiobook platform ACX estimates that about 9,300 English words equal one finished hour, roughly 155 words per minute. We know of no reliable published benchmark for Arabic voiceover pace, so always listen to the result.

02Why Arabic word counts differ from English

An Arabic word is usually heavier than an English one, because prepositions, pronouns and conjunctions attach to it. A single word such as وبمدرستكم means "and at your school", four English words. Count words and the Arabic script looks shorter, yet it takes the same time or longer to say.

• Attached pronouns and particles: فسأرسلها is one word; in English it is "so I will send it".
• MSA is usually slower: careful articulation plus case endings and nunation add syllables, hence 120 words per minute.
• Dialects move faster: everyday speech drops syllables, and Egyptian and Lebanese tend to run a little quicker than Saudi and Gulf.
• Numbers deceive: "2026" counts as one word but is spoken as four in Arabic, ألفين وستة وعشرين.
• Foreign brand names: a long English brand inside an Arabic sentence takes longer than the word count suggests.

There is a second yardstick that helps with budgeting: on Vocla, 1,000 characters is roughly one spoken minute, credits are counted by characters rather than words, and tashkeel marks are free. Know your character count and you know both duration and cost.

03Word counts for 15, 30 and 60-second spots, with examples

Examples make the numbers concrete. Here is a Saudi spot targeting 15 seconds at about 30 words, inside the 32-word limit at a natural pace:

And the same brief in English for 15 seconds, at about 36 words against a 38-word limit:

1. 115 seconds: one idea, one benefit, one call to action. Don't try to land three messages.
2. 230 seconds: a hook, two or three benefits, then the offer and the call to action.
3. 360 seconds: room for a short story or product explanation, broken into short sentences with clear pauses.

04Check your script with the length calculator

Instead of counting by hand, paste your script into the voiceover length calculator. Pick the language or dialect and the pace (calm, natural or lively) to get an estimated duration in seconds and how many words fit 15, 30 and 60-second slots:

The calculator counts words and sentences for you and adds a breath after each sentence. It is an estimate, so use it as step one: if it says your 30-second script runs 34 seconds, cut a sentence before you generate, not after.

05Pace changes everything: calm, natural, lively

One number does not suit every spot. A clinic or bank needs calm confidence; a White Friday sale needs energy. Here is how pace shifts the figures, using Saudi Arabic and English:

In Vocla's Text to speech editor you set pace with the speed slider and energy with the Boost option for punchy promos. But speed is not a fix for a long script: a voice racing at 1.3x to fit the slot sounds rushed, however good it is. Cutting words always beats speeding up.

06From word count to finished file in Vocla

1. 1Write and measure: draft the script, run it through the calculator, and trim anything over the slot.
2. 2Spell numbers out: write "twenty percent" rather than 20% so you know its real length and control the read.
3. 3Choose dialect and voice: pick from the dialect picker, then audition two voices on the first sentence.
4. 4Generate and check the duration: the player shows the real length, which beats any estimate.
5. 5Fix a sentence, not the file: if you are two seconds over, shorten one sentence and regenerate only that line.

Cost is easy to work out too: a 30-second spot is about 500 characters, roughly 600 credits on Vox Prime, 500 on Vox Stream or 400 on Vox Edge. Three versions in three dialects stay inside Starter's 5,000 credits for $4.99, with commercial rights.

Timing is one step in a longer process that starts with the brief and ends with a quality check. For the full path from script to final audio, read our voiceover workflow guide and our tips for writing scripts for AI voice.

07Common timing mistakes

• Translating the English script word for word: 76 English words fit 30 seconds, but a literal MSA translation may need 35 seconds or more.
• Forgetting the open and close: a sonic logo, web address or phone number eats precious seconds.
• Ignoring symbols: %, @ and URLs are all read out as words.
• Relying on speed: pushing the tempo to squeeze in a long script hurts clarity more than it helps.
• Not testing on a phone: the same duration feels faster on a small speaker, so test where your audience listens.

For TikTok and Reels, the first two seconds matter more than any calculation. A hook of four or five words is enough to stop the scroll; the rest follows. Our guide to TikTok hooks in Arabic dialects goes deeper.

08Long-form: YouTube, courses and podcasts

The same maths works for long-form content, but mistakes cost more: an extra minute per episode adds up to hours over a season. Rough figures at a natural pace, before paragraph pauses:

Split long scripts into short paragraphs and keep each generation under Vocla's 10,000-character limit, roughly ten spoken minutes, checking pronunciation before moving on to the next part.