20,000 characters: Sume TTS $0.95 vs ElevenLabs $0.80–$1.60

Turning a 20,000-character training module into speech costs $0.95 on Sume. ElevenLabs lists $0.80 for its Flash and Turbo models and $1.60 for its multilingual v2 and v3 models. Sume is above Flash by $0.15 and below the multilingual models by $0.65.

The arithmetic

The Sume public catalog lists Text to speech at $0.0475 per 1,000 characters; ElevenLabs' API pricing page lists $0.04 per 1,000 characters for Flash and Turbo and $0.08 for v2 Multilingual and v3 (both read 2026-10-09). Price is characters divided by 1,000, times the rate.

How to read it

The cheapest line is ElevenLabs Flash. Voices differ per service, so a lower price is only a saving if you accept that voice. Sume reserves the amount from the transcript character count and rounds the reservation up to a whole cent, so the reservation for a very short text is at least one cent.

If your volume is 20,000 characters a month, the monthly gap against ElevenLabs multilingual is $0.65; against Flash it is $0.15 the other way.