Voice Agent Cost Calculator
Estimate combined STT + LLM + TTS cost per call.
With the default inputs, $0.0690 per call — $0.0138/min · ~$1,035.00/month. Enter your own numbers below to recompute instantly; the full step-by-step math is shown under the worked example.
Results update automatically as you type.
- STT / min
- $0.00600
- LLM / min
- $0.000300
- TTS / min
- $0.00750
- Per call
- $0.0690
- Per month
- $1,035.00
A voice agent pays for transcription (STT), the language model, and speech synthesis (TTS) every minute. Combine all three to see the cost per call and per month.
How this is calculated
A voice agent stacks three costs per call: speech-to-text (billed per minute of audio), the language model (billed per token), and text-to-speech (billed per character). We compute STT = minutes per call × STT per-minute rate, LLM = minutes × tokens per minute × per-token rates, and TTS = minutes × characters per minute × TTS rate, then sum them for cost per call and multiply by calls per day × 30.
Which piece dominates depends on talk time versus token density: long calls with a premium model tip toward the LLM, while short, chatty calls are dominated by STT/TTS minutes.
Is this a good result? What to do next
Compare cost per call to the value of the call — a booking, a resolved support issue, a qualified lead. Voice-agent calls typically land in the single-digit-cents to low-dollars range per call; if you're near a dollar per call, check whether talk time or model choice is the cause.
Typical planning ranges
- Transcription (STT)
- ~$0.003–$0.006 / min
- Text-to-speech (TTS)
- ~$15 / 1M chars (~$0.015 / min)
- Typical 5-minute call (mini model)
- ~$0.05–$0.25
Ranges are typical planning figures to sanity-check your result, not authoritative benchmarks. Your numbers will vary with use case, volume, and vendor.
How to improve this number
- Shorten average call length — every minute bills all three components.
- Use a mini LLM and a cheaper STT/TTS tier where quality allows.
- Reduce TTS characters (concise responses) and LLM tokens per minute.
Common mistakes
- Costing only the LLM and forgetting STT and TTS, which often exceed it.
- Using per-token thinking for STT/TTS, which are billed per minute/character.
- Ignoring silence and hold time that still counts as billed minutes.
When to use a different approach
For a text-only assistant drop the voice layers and use the chatbot cost calculator. To size just transcription or synthesis, use the speech-to-text or text-to-speech calculators.
Worked example (defaults)
With the default inputs above, here is the result:
- STT / min
- $0.00600
- LLM / min
- $0.000300
- TTS / min
- $0.00750
- Per call
- $0.0690
- Per month
- $1,035.00
Sources & references
Frequently asked questions
What dominates voice agent cost?+
It depends on talk time. STT and TTS scale with minutes, while the LLM scales with tokens. Long calls with a premium model tip the balance toward the LLM.
Related calculators
Answering a real question?
This calculator powers these problem-solving guides: