Billing
Usage-based pricing — tokens, the orchestration fee, and the math.
Usage-based, no subscription
There are no plans or monthly fees. You keep a single token wallet (1 token = $1 of usage credit) and spend it as you go. Every account includes unlimited AI agents and one phone number for testing; add more numbers anytime and you just pay the telephony usage.
How a charge is calculated
Provider costs (AI, speech-to-text, text-to-speech, telephony) pass through AT COST from your wallet. The platform's only margin is a flat orchestration fee:
Total = Provider cost (at cost) + Orchestration fee Voice: $0.03 per minute Text: $0.003 per message
The orchestration fee covers call routing, the real-time runtime, logging, and analytics. There is no per-token markup — what the providers charge us is what you pay.
Worked example — one minute of voice
Using GPT-4o Mini + Deepgram + ElevenLabs Flash + Twilio (~1,500 LLM tokens and ~800 TTS chars per call minute):
| Component | Cost / min |
|---|---|
| Speech-to-Text (Deepgram) | $0.0048 |
| Text-to-Speech (ElevenLabs Flash, 800 chars) | $0.0400 |
| Telephony (Twilio inbound) | $0.0085 |
| LLM (GPT-4o Mini, ~1.5k tokens) | ≈ $0.0005 |
| Orchestration fee | $0.0300 |
| Total | ≈ $0.084 / min |
Tokens
Top up your wallet anytime — buy token packs ($25, $100, $500, or any custom amount) at $1.00 per token, always at cost.
- Purchased tokens never expire — there's nothing to reset each month.
- Turn on auto top-up to refill automatically when your balance runs low.
- There is no negative balance — when you hit zero, usage pauses until you add more.
- Bring your own provider keys (BYOK) per provider in Integrations and that usage spends no tokens — only the orchestration fee applies.
Enterprise
Running high call volume? Enterprise accounts get a negotiated (or waived) per-minute deal. Use the Contact us button on the pricing page or in Billing to reach out.