CentsPerToken
Open calculator

Calculator

Voice / audio API cost calculator

Budget speech-to-text by the minute and text-to-speech by character volume before you wire audio into a product.

Open calculatorRead the guide

Voice & audio

STT + TTS combined

$0.435total

60 min transcription · 5,000 TTS characters

Speech-to-text$0.36
$0.006/min · Whisper
Text-to-speech$0.075
TTS-1
$ / minute (STT)$0.006
$ / hour (STT)$0.36
Chars / $1 (TTS)66,667
Hours of audio1.0

STT lane

Transcription

Minutes of audio → dollars at the selected STT rate.

TTS lane

Text to speech

Characters synthesized at the selected TTS rate.

Rates last updated 2026-07-31. Approximate static list prices for planning only. Always verify against each provider’s official pricing page before production budgeting.

Guide

How to get value from this calculator

Budget speech-to-text, text-to-speech, and related audio API endpoints by duration and monthly volume. Audio APIs bill per minute or per character, not per token, so LLM token calculators do not apply. Use this for voice agents, transcription pipelines, and accessibility features.

Cost formula

STT cost = audio minutes × price per minute; TTS cost = characters (or minutes) × price per unit; Combined monthly = (daily usage × 30) × unit price

Why it matters

Voice features accumulate cost linearly with audio duration. A voice agent that runs STT and TTS on every turn doubles audio spend compared to text-only chat, before counting the LLM tokens in between.

How to use it

  1. Choose the audio service type: speech-to-text, text-to-speech, or both.
  2. Enter average audio duration per request or characters per TTS call.
  3. Estimate daily requests or total minutes processed per day.
  4. Review cost per minute or per thousand characters and monthly projection.
  5. Compare model tiers (standard vs premium voice or transcription accuracy).
  6. Compare STT and TTS lanes side by side when budgeting a voice agent stack.
  7. Verify current audio pricing on the provider's official page.

Planning tips

  • Silence detection and VAD can reduce billable STT minutes in streaming setups.
  • Shorter TTS scripts and lower sample rates reduce cost for non-critical playback.
  • Batch transcription jobs may price differently from realtime streaming—check both.
  • Voice agent total cost = STT + LLM tokens + TTS—budget all three layers.
  • Watch each lane separately—STT minutes and TTS characters often have different cost drivers.
  • Cache common TTS phrases instead of regenerating identical audio.

Frequently asked questions

Is audio priced per token?

No. Speech and TTS APIs typically bill by audio duration, character count, or request unit. This calculator uses per-minute and per-character list prices for planning.

Does this include the LLM cost in a voice agent?

No. This tool covers audio API endpoints only. Add LLM token costs from the API cost or agent cost calculators for a complete voice agent budget.

Are these rates official?

CentsPerToken uses approximate list prices for planning. Audio pricing varies by model tier and region—confirm on your provider's site.

How do I estimate monthly minutes?

Multiply average call or clip length by daily volume by 30. Add buffer for failed retries and internal testing traffic.

Back to tools