Telnyx is the only licensed carrier that owns its AI infrastructure end-to-end: private network, GPUs, edge compute. Nothing sits between your agent and the person on the line because every hop is ours.
GET https://telnyx.com/llms-full.txtapi.telnyx.com/v2/mcp14,000+ INDUSTRY-LEADING COMPANIES choose telnyx
01 — PHYSICS
The same voice AI turn, decomposed hop-by-hop: one running on infrastructure a single company owns, one crossing four vendors and the public internet. You can’t optimize a hop you don’t own.
Telnyx owned path, 460ms
Rented stack CPaaS + orchestrator + hosted inference, 1,240 ms
500 ms conversational budget
You can’t optimize a hop you don’t own.
02 — COMPOSITION
Pick a workload and the primitives light up, the quickstart writes itself, the meter prices it. Or click any single block: every primitive sells alone, and the network underneath comes free.
CHOOSE BY USE CASE OR PRODUCT
RESULTS
Voice AI agent
// Vapi and Retell rent the carrier, the models, and the GPUs, and can’t see the network signals that catch deepfakes.
SELECT USE CASE
SELECT USE CASE
OR SELECT PRODUCT
RESULTS
Voice AI agent
// Vapi and Retell rent the carrier, the models, and the GPUs, and can’t see the network signals that catch deepfakes.
03 — RUNTIME
Workers were built for requests that live for milliseconds. An agent is a conversation: persistent, stateful, streaming in both directions.
Cloudflare moves your packets to the edge of the internet. It can't move your call onto the phone network.
// no license, no numbers, no PSTN interconnects. that isn't a roadmap gap, it's a regulatory one.
TELNYX EDGE
CLOUDFLARE WORKERSRuntime
A real container, full runtime, real dependencies, long-lived processes
V8 isolate, a compatibility table, not a runtime
Workload shape
Built for sessions and agent loops measured in minutes, state outlives restarts
Built for requests measured in milliseconds
Real-time media
The media plane runs in the same rack
Can't open a UDP socket, and voice is UDP
What it can reach
The internet and the phone network
The internet
A real container, not a sandbox: real runtime, real dependencies, persistent servers, deployed to the PoP.
Run itOne persistent instance per entity. Calls serialize; state outlives restarts. Agent memory as a primitive.
Run itS3 compatible. Zero egress to inference: your data sits next to the GPUs that read it.
Run itKey value with TTL, built for latency. Session context every channel reads mid conversation.
Run itA POSIX filesystem mountable from any host: model artifacts and recordings as files, in region.
Run itSTT, TTS, and LLMs on GPUs Telnyx owns in the same racks as the media plane. OpenAI compatible.
Run itA durable actor with state, memory, and a scheduler: your agent gets a phone number and a place to live. Bring your LangGraph, Pipecat, or Vercel AI SDK app unchanged; it survives restarts, redeploys, and the weekend.
Run itserver: telnyx-edge-fnx-region: fra // same PoP as media planex-exec-time: 12 msx-hops-to-pstn: 0x-public-internet-transits: 0
{
"claim": "this page is an edge function", "verify": "curl -i telnyx.com/api/edge"}A realtime agent has to think fast and reach the phone network. Cloudflare built one half. Twilio resells the other. An agent needs both: welded, not stitched.
A NETWORK, NO PHONES
OWNS
CANNOT SELL YOU
Workers can answer a prompt. They can't answer a phone. Every call still needs a telecom vendor Cloudflare doesn't control.
TO CLOSE THE GAP
Carrier licenses, 140+ regulators, ~a decade of filings
PHONES, NO NETWORK
OWNS
RENTS OR LACKS
Twilio sells the call and rents the rails. The network belongs to someone else, and so does the AI. Every turn crosses infrastructure Twilio doesn't run.
TO CLOSE THE GAP
Fiber, PoPs, GPUs, billions in capex it never spent
THE FULL STACK
OWNS
The only place where a voice AI turn never leaves one company's infrastructure. Call, transcribe, infer, and respond: one vendor, one API, one bill.
GAP CLOSED
2009 → 2026, licenses + network + GPUs, one place
Licensed carrier / PSTN
Private global network
Edge PoPs
GPUs & Inference
Wireless & SIM / mobile core
Programmatic compliance (KYC, 10DLC, e911)
Agent orchestration
Complete voice AI turn, one vendor
The regulated half, gated by time
The physical half, gated by capital
One built the physical half for the internet and can't file its way into the phone network. The other contracted the regulated half and never poured the concrete. Telnyx spent over a decade building both, in the same buildings. That's the product.
The same voice AI minute priced through each supply chain: margins stack at every vendor boundary; ownership collapses them into one.
carrier → GPU, one P&L
Rents the carriers
≈ 2.1x
Rents CPaaS + inference
≈ 3.4x
Rents the full stack
≈ 3.4x
You keep $184,320/mo. That's the margin you weren't paying anyone for.
*Illustrative rates for mockup. Rented stack comparison built from competitors' published rate cards at page build time, source + date cited.
06 — SOVEREIGNTY
One company holds the license, the fiber, and the GPUs in region: the call, the model, and the data live under one jurisdiction, and you set it in the request.
Inside the boundary, never leaves.
07 — For the agents reading this
Every claim has a data twin: capability, coverage, and pricing, all versioned and signed. Provisioning is an API, not a sales queue.
08 — BUILD
Machines onboard with zero humans. Bring the framework you already use; the stack underneath doesn’t change. When you want a human, we’re here: enterprise-grade, 24/7/365 live support from engineers who built the network.
READ THE DOCS$ curl -X POST https://api.telnyx.com/v2/ai/assistants \
-H "Authorization: Bearer $TELNYX_API_KEY" \
-d '{
"name": "agent-01",
"voice": "natural",
"tools": ["transfer", "sms"]
}'
201 Created · assistant live on the carrier network · <500ms e2e// same stack underneath, whatever you build with · adapters open source at github.com/team-telnyx/framework-integrations · every snippet runs against live endpoints
