Text-to-speech API free or unlimited? What $1 buys on Sume

"Text to speech API free unlimited" is on Google's autocomplete (read 2026-10-07). Sume's catalog rate card, the table every audio job is priced from, lists a metered price for each audio product and no free or unlimited line. This page is the practical answer: what one dollar of balance buys on each.

All prices are the provider list times the 1.25 house margin, taken from the repo's catalog on 2026-10-07. GET /v1/catalog returns the live values for your account.

One dollar of audio

Two rounding rules matter at small sizes. Each TTS job is rounded up to the next cent, so a 40-character line costs $0.01, not $0.0019, and a hundred of them cost $1.00. Lip-sync and STT are counted by whole seconds and minutes.

Make a small test cost almost nothing

Use dry_run to price a job without running it and max_spend_usd to cap one that does run. A TTS request is cheap to retry because it takes an Idempotency-Key: the same key returns the same job instead of billing twice.

curl -X POST https://api.sume.com/v1/tts-1.0/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: tiny-test-001" \
  -d '{"transcript": "Hello from Sume.", "voice": {"id": "YOUR_VOICE_ID"}}'

Reading the table

Speech is the cheap part of an audio project: a 3,000-character script is 15 cents. The expensive lines are the ones that make pictures move, so a budget for a voiced video is mostly lip-sync seconds. Plan the script first, price the speech in cents, then spend what is left on the clips.