$0
Standard rates. Live in minutes, card on file.
- Full platform, every product
- Voice engine $0.05/min, LLM at GPU token rates
- 500 concurrent calls, 100 API req/s, 50 SMS/s
- Community support
VOICE AI PRICING
$0.05 per minute. No fine print.
INCLUDEDVoice engine
$0.05Per minute
OrchestrationFull agent runtime - turn-taking, interruption handling, tools, knowledge base retrieval.
Text-to-speechTelnyx Ultra, Natural & NaturalHD, Qwen3TTS, Inworld, Rime, Resemble, and Murf.
Speech-to-textDeepgram, Telnyx STT - Telnyx-hosted, real time.
ADD-ONLLM tokens
On Telnyx owned GPUsKimi, typically ~$0.006/min, no prompt-size multipliers.
Or on a managed frontier modelAnthropic & OpenAI models rates.
ADD-ONTelephony
Carrier costStarting at $0.0032/min.
One invoiceOne Telnyx invoice instead of five different vendor invoices.
ESTIMATE YOUR COST
Your numbers, itemized
Set your volume and call profile. Every line of the estimate maps to a line on your invoice, and we'll show what the same month runs elsewhere.
ADJUST SETTINGS
0500K
Call profile
LLM
*Telnyx estimates exclude per-call voice-engine rounding to 60-second increments, so actual voice-engine charges can be higher, especially for short calls. LLM usage is billed per token; telephony follows its own billing rules. LLM per-minute figures are estimates at roughly 6,000 prompt tokens per minute for a realistic production agent with tools, knowledge retrieval, and transcript, with prompt caching and 500 output tokens. Actual billing is per token at the rates in the tables below: no prompt-size multipliers, no minimums, and caching means big prompts cost far less than list rate. Competitor estimates from public pricing and independently reported production costs, July 2026.
PLANS
Pick your commitment, not your features.
Every plan runs the same platform, same models, same network. Commit more, pay less per minute.
$500Monthly minimum
Volume discounts on everything you use.
- Everything in pay as you go plan
- Volume discounts across all usage
- Dedicated account manager
- Priority support and higher rate limits
$5,000Monthly minimum
Custom pricing on dedicated infrastructure.
- Everything in committed plan
- Dedicated infrastructure and IP addressing
- Private network interconnect
- Unlimited rate limits
PAY AS YOU GO
Building with components instead?
The Voice AI voice engine bundles orchestration, hosted STT, and hosted TTS at $0.05/min, with LLM tokens at the rates below. If you're using pieces standalone, STT for transcription, TTS in your own app, inference over API, these rates apply.
| COMPONENT | RATE |
|---|---|
| Voice API platform fee | $0.002 / minSIP trunking billed on top |
| SIP trunking — inbound | from $0.0032 / min |
| SIP trunking — outbound | from $0.005 / min |
| Call recording | $0.002 / min |
| Media streaming (WebSockets) | $0.0035 / min |
| Conference | $0.002 / participant / min |
| Phone numbers (US local) | $1.00 / moVolume pricing from $0.79/mo |
| Number lookup | from $0.0015 / query |
| SMS outbound | from $0.004 / msgVolume tiers down to $0.0005/msg |
| Fax (send or receive) | $0.007 / page |
COMPETITION
Built for agents. Not retrofitted for them.
On most platforms, agents sit on top of UI-first products, constrained by rented infrastructure and manual approval paths. Telnyx is different. We're the only agent-native platform that owns the primitives, not abstracting over someone else's foundations.
FEATURE
Voice engine rate
$0.05/min
$0.05/min
$0.055/min
$0.08/min
Orchestration, STT, TTS included in sticker
LLM billing
Per token on Telnyx GPUs
Pass-through, per token
$0.003–$0.08/min listed
Passed through, per token
Realistic all-in
≈ $0.06/min
$0.12–$0.42/min
$0.088–$0.31/min
$0.10+/min
No monthly platform fee
Concurrency included
500 calls
10 lines, + $10/line/mo after
20 calls
4-40 by plan, then 2x burst pricing
HIPAA included in every plan
Owns the network
Owns AI inference
Compare what 10K minutes could cost you per month
$596per month
$1,450-$3,250*per month
$880-$1,650*per month
$880-$1,650*per month
Telnyx estimates exclude per-call voice-engine rounding to 60-second increments, so actual voice-engine charges can be higher, especially for short calls. LLM usage is billed per token; telephony follows its own billing rules. * Competitor estimates from public pricing and independently reported production costs, July 2026.
COMPLIANCE
Compliance is included. Not an add-on.
Some voice AI platforms charge up to $2,000/month extra for HIPAA. On Telnyx, regulated-industry readiness is standard on every account, on every plan, including pay as you go.
