Voice AI API pricing, per engine and per second
$0.0491 to $0.1851 per connected minute, billed per second from one prepaid balance. No subscription, $5 to start.
Pay as you go
$0.0491 to $0.1851 per connected minute, by engine
- Every voice engine and its models, on one API key
- Your own carrier, billed by the carrier, with no markup
- Apps on a call through MCP, with no fee per tool call
- Billed per second from one prepaid balance
- $5 minimum top-up, no subscription
- $0.012 per minute platform fee, already in every engine price
- 100,000 API requests a month included
- 5 calls at once to start, more on request at no charge
- Every charge itemised in usage, to the second
What a call costs on each engine
Per connected minute, platform fee included. A 3-minute call is the price times three, billed per second.
| Engine | Per minute | 3-min call |
|---|---|---|
| ClassicDeepgram + GPT-6 Luna + ElevenLabs | $0.0865 | $0.2595 |
| SarvamSarvam, Indian languages end to end | $0.0617 | $0.1851 |
| Gemini LiveGemini Live, native audio | $0.0491 | $0.1473 |
| Deepgram AgentDeepgram Voice Agent | $0.1851 | $0.5553 |
| OpenAI GPT-LiveOpenAI GPT-Live + GPT-6 Luna | $0.0827 | $0.2481 |
| OpenAI RealtimeOpenAI Realtime, gpt-realtime-2.1-mini | $0.0850 | $0.2550 |
| Grok VoicexAI Grok Voice | $0.1240 | $0.3720 |
Telephony is not included and not marked up. Phone calls run over your own SIP trunk, so you keep your carrier rates and your carrier relationship. Web and app sessions are billed at the same per-minute rates, with no telephony at all.
Build a pipeline
Pick a speech-to-text model, a language model and a voice. The price is what the three add up to per minute.
Per connected minute
Runs on the Deepgram · OpenRouter · ElevenLabs pipeline, which the speech-to-text selects. At typical speech rates, including the platform fee; metered per second, token and character, and billed for the model that actually ran. Within a family the pipeline picks the variant for the agent’s language. How composing works
Paste into POST /v1/agents
{
"voice": {
"stt": {
"model": "deepgram/nova-3"
},
"llm": {
"model": "openrouter/openai/gpt-6-luna"
},
"tts": {
"model": "elevenlabs/eleven_flash_v2_5"
}
}
}What else is charged?
Four things, all from the same prepaid balance and all itemised in usage. Everything else an account does costs nothing.
| Charge | Price | When |
|---|---|---|
| A cloned voice | $1.00 | Once, when the clone is made. Speaking in it costs the same as any other voice. |
| A WhatsApp message | $0.002 | Each message an agent sends or receives, plus that agent's usage at the rates above. Meta bills its own WhatsApp fees to the business directly. |
| API requests past 100,000 a month | $0.00001 | Per request, once the month is over. Requests that are rate limited or rejected are not counted. Charging starts with November 2026. |
| Platform work outside a call | Provider cost | Post-call analysis, building an agent or a mascot, importing knowledge, reading a scanned PDF, indexing a knowledge file, making a voice preview. Charged at the same per-token and per-character rates as a call, with no orchestration fee. |
Platform work is fractions of a cent at a time: a knowledge import is typically well under a cent. It is billed rather than absorbed because it is a real provider call. Every charge, line by line.
Pricing questions
What does Wixzel Voice cost per minute?
From $0.0491 (gemini-live) to $0.1851 (deepgram-agent) per connected minute at typical speech rates, including the flat $0.012 per minute orchestration fee. Every figure is generated from the price book the meter bills against and rounded up, never down. Telephony runs over your own SIP trunk at your carrier's rate and is not included or marked up.
Is there a subscription, a seat price or a minimum monthly spend?
No. Credit is prepaid from a $5 minimum top-up and spent per second of actual usage. An account with no calls spends nothing.
What is charged besides the per-minute rate?
Four things. A cloned voice costs $1.00 once, when it is made. A WhatsApp message costs $0.002 each way, and Meta bills its own WhatsApp fees to the business directly. API requests past 100,000 a month cost $0.00001 each. Work the platform does outside a call, such as post-call analysis, building an agent, importing knowledge or making a voice preview, is charged for the provider usage it costs at the same rates as a call, with no orchestration fee: fractions of a cent at a time. Everything comes from the same prepaid balance and every charge is itemised in usage.
Are there limits on how many API requests I can make?
An account includes 100,000 API requests a month. Going over is allowed, because it is a price and not a cap: each further request costs $0.00001, charged once the month is over. A request that is rate limited or rejected is never counted and never charged. Charging starts with November 2026, so no earlier month is billed. The separate per-second rate limit is 10 requests a second with a burst of 20.
How are web and app sessions billed?
Exactly like a phone call on the same engine: the same per-minute price, from the same balance, per second. A session counts toward your concurrent-call limit, appears in your call log with
channel: "web" and ends with the same callCompleted webhook. There is no telephony cost, because there is no phone line.How am I billed?
Prepaid. You add credit, and calls spend it per second of speech, per token and per character. A call in progress is allowed to finish, so a balance can end a little below zero, by at most $0.50. The minimum top-up is $5.
What happens when credit runs out?
A call that cannot fund its first 15 seconds is refused with a 402 before anything is allocated. Mid-call, at about a minute of runway, the agent warns your caller, then speaks a closing line and hangs up. It does not drop the line mid-sentence.
Can I retry a request safely?
Send an
Idempotency-Key. It is required on POST /v1/calls, POST /v1/agents/{id}/test-call, POST /v1/phone-numbers/purchase and POST /v1/billing/topups, and a replay with the same key and body returns the original response instead of dialling twice. Keys are kept for 24 hours. POST /v1/campaigns/{id}/start takes no key, and starting a stopped campaign dials every lead on it again.Is there a free trial?
There is no time-limited trial and no free tier. Credit is prepaid from $5, a short test call costs a few cents, and the Credits and Refunds policy explains what can be refunded. You can talk to each engine from the landing page, and each new account gets 5 free minutes to talk to its own agents from the console, before adding any credit.
How many calls can run at once?
Five per account by default, phone calls and web sessions counted together. Past the limit the next call is refused with a clear error and the calls already running are untouched. The cap is raised on request; email the address on the contact page with the volume you expect.
Build your first voice agent
Add $5 of credit to start. No subscription, billed per second.