Annual compliance refresher voiceover in six languages: what it costs
A 3,000-character training script in six languages costs $0.855 on Sume TTS. Set language per voice, translate yourself, and re-run only what changed.

A 3,000-character refresher script voiced in six languages is 18,000 characters in total. At Sume's $0.0475 per 1,000 characters that is $0.855, or $0.1425 per language (API reference). Translation is not included: Sume TTS reads the text you send and, per its contract, never translates a non-English request into English. You supply the six scripts.
The six-language budget
Costs follow the character count of each translated script, not the English one. Languages written in fewer characters come out cheaper, and ones with longer words cost more, so recompute with the real text.
| Language | Code | Characters (assumed) | Cost |
|---|---|---|---|
| English | en | 3,000 | $0.1425 |
| Spanish | es | 3,000 | $0.1425 |
| French | fr | 3,000 | $0.1425 |
| German | de | 3,000 | $0.1425 |
| Japanese | ja | 3,000 | $0.1425 |
| Korean | ko | 3,000 | $0.1425 |
| Total | 18,000 | $0.855 |
Set the language on every non-English job
language is required for non-English text. Omitting it defaults to English at the provider, and the API infers Korean or Japanese only as a fallback when the transcript is Hangul or kana alone. Each language needs a voice that speaks it. If the voice and language disagree, the job fails with tts_voice_language_warning unless you confirm the mismatch on purpose.
Check the audio limits
One request takes up to 20,000 characters and audio up to 1,200 seconds. A 3,000-character script is well inside both. For a 20-minute course, split by chapter and join the parts with timeline audio, a flat $0.01 per join of up to 20 parts (timeline audio docs).
Why this is a live question
A single Microsoft voice is now described as speaking 23 languages with native accents, at $22 per million characters, with a Flash tier at $15 (Microsoft AI, read 2026-10-04). For a compliance team the price gap is small next to the cost of a wrong translation. Spend your time on review, not on the voice.
Re-run only what changed
Compliance text changes by a clause a year. Name each job key after language and section (ko-s3-2026), re-render only the changed sections and rejoin. Last year's six-language set then costs you a few cents to update, not $0.855 again. See the eight-language fan-out guide for the loop.
Sources
Related posts
More in Use cases
- App preview clip from screenshots with Seedance 2.5 and a caption pass
Screenshots as reference images can drive a 15 s Seedance 2.5 app promo on Sume for $8.67 at 720p. Generated UI text is unreliable, so add words after.
- Arabic AI video: Luma's Arabic UI vs Sume prompts
Luma added an Arabic interface across Luma Agents. Sume docs list ko-KR chrome but no Arabic UI, so verify Arabic speech and text output yourself.
- Article 50(2): the provider duty to mark AI outputs machine-readably
Article 50(2) makes providers mark synthetic outputs in machine-readable form. What the wording says, what is exempt, and what to ask a generator vendor.
- Article 50(2) assistive editing: trim and crop vs generate
Article 50(2) exempts assistive editing that does not substantially alter input. Sort your Sume outputs into generated and file-processed before you rely on it.
Written by Sume