Narrate an 80,000-word novel with AI TTS: $23 in 34 requests

An 80,000-word novel is about 480,000 characters, quoted at $23.00 on Sume TTS in 34 requests of up to 14,400 characters, roughly 10 hours of audio.

5 min readSume
All posts

An 80,000-word novel is about 480,000 characters at 6 characters a word, which Sume text to speech quotes at $23.00 in total when split into 34 requests, and roughly 10 hours of audio at 750 to 800 characters a minute. Per-request rounding adds almost nothing: the raw price is $22.80. The assumptions are mine: 6 characters a word and parts of 14,400 characters, chosen to stay under the 1,200-second audio limit per request.

Rate and limits are from the Sume pricing and API docs; the pace is Cartesia's published rule of thumb of 750 to 800 characters per minute of audio (pricing page, read 2026-10-03). Treat the hours as an estimate; the real pace depends on the voice.

The arithmetic

480,000 characters divided by 14,400 is 33.3, so 33 full parts and a final part of 4,800 characters (800 words). A full part is 14,400 x $0.0475 / 1,000 = $0.684, rounded up to $0.69. The last part is $0.228, rounded up to $0.23.

Novel narration quote (as of 2026-10-03)
ItemPartsCharacters eachQuote eachSubtotal
Full parts3314,400$0.69$22.77
Final part14,800$0.23$0.23
Total34480,000$23.00

Length and chapters

At 750 characters a minute the novel is 640 minutes, 10 hours 40 minutes; at 800 it is 600 minutes, 10 hours. Each full part runs 18 to 19.2 minutes, safely under the 20-minute cap. Real chapters rarely land on 14,400 characters, so split on paragraph breaks near that size and keep the cap in mind: a 3,000-word chapter at 18,000 characters would be 22.5 to 24 minutes and would fail.

If chapter files should be a single audio per chapter, join the parts. Timeline audio concat takes 1 to 20 parts; see Timeline audio. Joining the 34 parts takes two concat jobs of up to 20 parts each, or one job per chapter.

What Sume does and does not do

Sume text to speech is $0.0475 per 1,000 characters, each quote rounded up to whole cents, up to 20,000 characters or 1,200 seconds of audio per request. Speech to text is $0.01 per minute of audio, up to 600 seconds per job.

Every request uses one voice. It does not read a manuscript file, detect chapters or assemble an audiobook in a store-ready format; your code does the splitting and the final upload. Marketplace-specific rules are outside this page; I did not verify any retailer's requirements.

The existing walk-throughs cover the mechanics: long text and audiobooks and the 1,200-second limit.

  • Count characters, not words, before you budget.
  • Cut at paragraph breaks and keep each part under about 15,000 characters.
  • Spot-check one part before running all 34.

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume