How many characters is a 15, 30, 45 or 60 second AI narration

Using 750 characters per minute of speech, a 15-second voiceover is 188 characters, 30 seconds is 375, 60 seconds is 750. Sume TTS cost for each.

4 min readSume
All posts

Text-to-speech is billed in characters, but video is planned in seconds. Unreal Speech's pricing page says one minute of audio is roughly 750 characters (read 2026-10-06). That is 12.5 characters per second, and it is a rule of thumb, not a Sume constant: a slow, dramatic read can use half that, and a fast ad read more. Measure a sample from your own voice and use that rate.

Seconds to characters to cost

Characters at 750 per minute and Sume TTS cost at $0.0475 per 1,000 characters rounded up to the cent; the 750 per minute figure is Unreal Speech's rule of thumb (read 2026-10-06)
Narration lengthCharactersSume cost per jobPer 1,000 characters, unrounded
15 seconds188$0.01$0.00893
30 seconds375$0.02$0.01781
45 seconds562$0.03$0.02669
60 seconds750$0.04$0.03563
90 seconds1125$0.06$0.05344
180 seconds2250$0.11$0.10688

Reading the table

  • A 15-second line is about 188 characters and rounds up to $0.01.
  • A 60-second script, 750 characters, is $0.04. The exact figure is $0.03563 before the per-job round-up.
  • A 3-minute narration, about 2,250 characters, is $0.11 in one job.

Count, then generate

In Python, len(text) counts characters the way the request does (Unicode code points, with a 20,000-character cap). Count the final script, not the draft, and subtract markup you will not send. Korean and Japanese scripts pack more speech into each character than English does, so the 750 rule overshoots for them; always test your own language.

len("Your approved narration goes here.")

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume