A 45,000-character chapter needs three Sume TTS jobs: about $2.14
Sume TTS caps a job at 20,000 characters. A 45,000 character chapter splits 20,000 + 20,000 + 5,000 and costs 45 x $0.0475, about $2.14.

Sume Text to speech accepts at most 20,000 characters per job, so a 45,000 character chapter becomes three jobs of 20,000, 20,000 and 5,000 characters. At $0.0475 per 1,000 characters the total is 45 x $0.0475 = $2.1375, which rounds to about $2.14.
The jobs
The rate and limit come from the API pricing page: priced per transcript character after Sume margin, spaces and punctuation counted, maximum 20,000 characters. The cap is per job, so splitting does not change the per-character price.
| Job | Characters | Cost at $0.0475 per 1,000 |
|---|---|---|
| 1 | 20,000 | $0.95 |
| 2 | 20,000 | $0.95 |
| 3 | 5,000 | $0.2375 |
| Total | 45,000 | $2.1375 |
Where to cut
Cut at a paragraph or sentence end, never mid-sentence, because each job produces its own audio file with its own pacing. The MCP tools page describes one tts_create per sentence for agent work, which is the other extreme: more jobs, smaller retakes.
A middle path is one job per scene of about 3,000 to 5,000 characters. Ten such jobs at 4,500 characters cost 10 x $0.21375 = $2.1375, the same total, but you can redo one scene for about 21 cents instead of redoing 20,000 characters for 95 cents.
Joining the parts
After synthesis, join the files rather than re-synthesising. Timeline audio concatenation is $0.01 per job, so joining three parts adds one cent. A whole 12-chapter book at 45,000 characters each is 540,000 characters, 12 x $2.1375 = $25.65 for the voice, before retakes.
Checking the character count
The per-job limit is on characters including spaces and punctuation. Count characters from the final text, not from a word processor's character-without-spaces figure. A chapter that shows 38,000 characters without spaces may be 45,000 with spaces and punctuation, which is the difference between two jobs and three.
Keep a small margin under 20,000 per job, for example 19,500, so a late edit does not push a job over the cap. With a margin of 500, a 45,000 character chapter still needs three jobs.
- Per-job cap: 20,000 characters.
- Rate: $0.0475 per 1,000 characters.
- Minimum charge per job: 1 cent.
Cost of retakes at each granularity
If one 5,000 character section needs a retake, redoing the 20,000 character job costs $0.95, while redoing a one-scene job costs about $0.24. Over a 12 chapter book with one retake per chapter, the extra cost is $11.40 with 20,000 character jobs and $2.85 with 5,000 character jobs.
That is why the finer split can be the cheaper plan even though the first-pass price is identical.
Sources
Related posts
More in Pricing
- Drop to 480p: Fabric saves 46.7 percent, H3 lip-sync 37.5 percent
Fabric is $0.10 at 480p against $0.1875 at 720p. H3 Max lip-sync is $0.0625 at 480p against $0.10 at 768p. Savings on 250 clips inside.
- 5 depositions of 3 hours: 15 hours of transcript, priced
Five 3-hour recordings are 15 hours: $8.10 on MAI-Transcribe-2-Streaming's intro price and $9.00 on Sume STT, in 90 jobs. Review needs are noted.
- 5-minute talking still on Fabric: $56.25 at 720p, $30.00 at 480p
VEED Fabric 1.0 on Sume accepts up to 300 seconds of audio: 300 x $0.1875 is $56.25 at 720p and 300 x $0.10 is $30.00 at 480p.
- $50 batch of 10-second 9:16 clips: 8 to 40 at 720p, by model
At 720p, $50 buys 8 ten-second vertical clips on Seedance 2.5 or 40 on Wan 3.0 and Gemini Omni Flash 1.1. Full table with clip price and change left over.
Written by Sume