All capabilities

Transparent per-minute cost. No black box.

See the exact per-minute cost of every call before you launch, and keep messaging spend predictable with per-conversation billing, daily caps and alerts.

Voice per minute · messaging per conversation

Voice · cost per call minute

$0.06

  • STT · Soniox$0.005
  • LLM · Gemini 2.5 Flash$0.020
  • TTS · Smallest Lightning$0.035
Total$0.06

Chat & messaging

Billed by tokens · one line per conversation · illustrative usage

  • WhatsApp≈ 3.6k tokens / reply~8 replies
  • Web chat≈ 3.6k tokens / reply~8 replies
  • SMS≈ 3.1k tokens / reply~5 replies
  • Email≈ 4.9k tokens / reply~3 replies
Daily spend capalert at 80%
  • Cached instructions billed at the lower cached rate
  • Your team's inbox replies carry no platform fee
  • Test conversations never count toward live budgets
Estimated cost per call minute~$0.14/min
  • STT — Deepgram Nova-2$0.012/min
  • LLM — Gemini 2.5 Flash$0.020/min
  • TTS — ElevenLabs Flash v2.5$0.10/min
  • Advanced Noise Cancellation$0.004/min
  • Post-call transcription$0.007/min
Total (live stack)$0.14/min

Example stack: Deepgram STT + Gemini 2.5 Flash + ElevenLabs Flash v2.5. First 5,000 prompt tokens included; overage priced clearly.

What you get

  • Cost broken down by component: STT, LLM, TTS, noise cancellation and transcription.
  • Shown in dollars, before a single call goes out.
  • Pick cheaper models to spend less, premium models when it matters.
  • First 5,000 prompt tokens included; overage is priced clearly.
  • No surprise analytics fees — observability is part of the platform.
  • Messaging is billed per conversation, not per message — each AI attempt billed once, never twice.
  • Daily spend caps in your workspace's time zone, with an alert before the cap; at the cap the AI pauses and hands conversations to your team.
  • Prepaid wallets and test conversations that never count toward live budgets.
  • Per-contact and per-workspace reply limits stop runaway costs from loops or abuse.

→ Know your cost per outcome up front — on calls and messaging — then tune the stack to protect your margin.

Run this on a real call.

Book a demo and see it live, end-to-end.

Book a demo