Start free with $10 in credits. No commitments, no contracts — scale up or down at any time.

Pay as you go

$0.00 to start

$10 Free credits included
  • Full platform access & templates
  • Call analytics and transcripts
  • Simulation testing & webhooks
  • Role-based access control & SSO
  • 20 free concurrent calls
  • 10 free knowledge base
  • Community & email support
Start for Free

Enterprise

Custom Pricing

  • Everything in Pay as you go
  • Dedicated stable server
  • No cap on concurrent calls
  • Developer API & SDKs
  • HIPAA / BAA compliance
  • Custom MSA & DPA terms
  • Named account manager
  • 24/7 priority support portal
Contact Sales

Calculator

Adjust the sliders to match your usage — see your estimate update in real time.

1. Talktime per month$0.408 / min total
minutes / month
2. LLM$0.320 / min
3. Speech-to-Text$0.0080 / min
4. Text-to-Speech$0.015 / min
5. Telephony$0.015 / min
6. Platform fee$0.050 / min
7. Add-ons
Batch Call$0.005 / dial
Branded Call$0.010 / outbound call
Safety Guardrails$0.005 / min
PII Removal$0.010 / min
AI Quality Assurance$0.010 after 100 min free
SMS$0.100 / message
8. Monthly subscriptions
Phone numbers$2 / month
Reserved Concurrency$8 / slot / month
Knowledge Bases (beyond 10 free)$8 / KB / month
Enterprise Support$1000 / month
White Label License$1000 / month
On-Prem Maintenance$3000 / month
Estimated Total
$408.00
/month
Talktime$408.00
$0.408 / min · 1,000 min

Voices
PlatformPlatform voices$0.015/min
Minimax voices$0.015/min
Fish voices$0.015/min
Elevenlabs voices$0.040/min
Cartesia logoCartesia voices$0.015/min
OpenAI voices$0.015/min
Deepgram Flux (Multilingual)$0.007/min
Deepgram Nova (Multilingual)$0.005/min
Deepgram Medical$0.065/min
OpenAI Whisper$0.008/min
OpenAI 4o Transcribe$0.006/min
Cartesia logoCartesia Ink 2$0.007/min
AssemblyAI logoAssembly Universal 3.5$0.0075/min
Azure AI Speech$0.008/min
GPT Real-time 1.5$0.345/min
GPT Real-time 2.0$0.38/min
GPT Real-time 2.1$0.38/min
GPT Real-time 2.1 mini$0.07/min
Gemini 2.5 Flash Native$0.20/min
LLM
GPT-5.5$0.32/min
GPT-5.4$0.16/min
GPT-5.4 mini$0.072/min
GPT-5.4 nano$0.01/min
GPT-5.2$0.112/min
GPT-5.1$0.008/min
GPT-5$0.008/min
GPT-5 mini$0.024/min
GPT-5 nano$0.006/min
GPT-4.1$0.0675/min
GPT-4.1 mini$0.024/min
GPT-4.1 nano$0.006/min
Claude 4.6 Sonnet$0.08/min
Claude 4.5 Sonnet$0.008/min
Claude 4.5 Haiku$0.025/min
Gemini 3.5 Flash$0.081/min
Gemini Flash 2.5$0.035/min
Gemini Flash 2.5 Lite$0.006/min
Gemini 3.0 Flash$0.027/min
Telephony
Persistence Twilio$0.015/min
Telnyx US Voice$0.03/min
Customer SIP Trunk (BYOC)Free
🇺🇸US$0.015/min
🇺🇸US (Toll free)$0.06/min
🇮🇳India$0.015/min
🇦🇺Australia$0.10/min
🇩🇪Germany$0.10/min
🇬🇧United Kingdom$0.10/min
🇫🇷France$0.10/min
🇧🇷Brazil$0.10/min
🇨🇦Canada$0.015/min
🇸🇬Singapore$0.10/min
🇯🇵Japan$0.10/min
🇰🇷South Korea$0.10/min
🇦🇪UAE$0.10/min
🇳🇬Nigeria$0.15/min
🇿🇦South Africa$0.15/min
🇲🇽Mexico$0.10/min
Add-Ons
Knowledge Base$8/KB/mo
Batch Call$0.005/dial
Branded Call+$0.010/call
Safety Guardrails+$0.005/min
Advanced Denoising+$0.005/min
PII Removal+$0.01/min
AI Quality Assurance$0.010/min
SMS$0.10/msg
Persistence Phone Number$2/mo
Custom Phone (BYOC)Free
Reserved Concurrent Calls$8/slot/mo
Community SupportIncluded
Enterprise Support$1,000/mo
White Label License$1,000/mo
On-Prem Maintenance$3,000/mo
  • ·Calls are tracked to the nearest second — no per-call rounding.
  • ·Agent fees stop the moment a call is transferred; telephony continues until the call ends.
  • ·Silence and hold time is billed — the STT engine remains active.
  • ·Failed or unconnected calls are never charged.

Get in touch

Talk to our team

Have questions about pricing, enterprise plans, or need a custom quote? Book a 30-minute call and we'll walk you through it.

  • Volume discounts for 10,000+ minutes/mo
  • Custom MSA & DPA for enterprise
  • HIPAA / BAA compliance
  • Dedicated account manager

Pricing questions

What teams ask before working out whether the numbers stack up.

What actually drives the cost of a voice agent?

Call minutes are the main driver, but they are not the whole picture. Longer calls cost more than short ones, so an agent that resolves quickly is cheaper per outcome even at the same rate. Telephony carries its own cost, and calls that escalate to a person still consume staff time. The useful number is cost per resolved call, not cost per minute.

How do I compare this against what we spend today?

Start from what a call currently costs you end to end — staff time, queue abandonment, and the calls nobody answers after hours. Then decide which specific intents an agent would take. Comparing a per-minute rate against a salary in the abstract tends to produce a number nobody believes; comparing cost per resolved call for one intent is defensible.

Is it cheaper to build this ourselves?

Sometimes, and it is worth being honest about when. Assembling speech-to-text, a model, text-to-speech and telephony yourself is achievable; keeping latency low, handling interruptions, and maintaining it as models change is the part that consumes engineering time indefinitely. The question is not whether you can build it but whether that is where you want your engineers.

Can we start small rather than moving everything at once?

That is the usual pattern and the one we would recommend. Move a single intent — after-hours answering, order status, appointment booking — and measure it properly before expanding. It keeps the blast radius small while you learn how your callers actually behave, which is rarely how anyone predicted.