A 100,000-character script: five Sume TTS jobs at $0.95 vs MAI-Voice

100,000 characters is five full 20,000-character Sume TTS requests at $0.95 each, $4.75 total; MAI-Voice-2.1 lists $2.20 and Flash $1.50.

4 min readSume
All posts

A 100,000-character script (about an audiobook chapter) takes five Sume TTS requests at the 20,000-character maximum, $0.95 each, so $4.75 in total. The same text lists at $2.20 on MAI-Voice-2.1 and $1.50 on MAI-Voice-2.1-Flash.

Cost of a 100,000-character script (read 2026-10-04)
RouteRequestsTotal
Sume TTS (20,000 chars max per request)5$4.75
MAI-Voice-2.1 ($22/1M)n/a$2.20
MAI-Voice-2.1-Flash ($15/1M)n/a$1.50

How to split the script

Cut at paragraph boundaries so no chunk goes over 20,000 characters, run the chunks as separate requests, and join the audio files afterwards, or place them on a Timeline. Joining on a Timeline is billed separately per output minute.

Check the price before you run

Sume lists the rate on the rate card; Microsoft's rates are in its announcement. The Sume maximum for a single request is $0.95, which is why five chunks is $4.75.

Recompute it in Python

Ceiling division gives the request count; each full request is $0.95. The numbers are the ones in the table above.

chars = 100_000
requests = -(-chars // 20_000)
print(requests, 'requests', round(requests * 0.95, 2), 'USD')

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume