GA Cartesia

Cartesia Managed Agents (Line)

Managed voice agents combining Cartesia Ink-2 speech recognition, an LLM you choose and Sonic-3.6 TTS, with phone numbers, SIP and evals. Suits teams that like Sonic's low-latency voices and want a managed agent runtime.

Est. per minute$0.06 - 0.09
1 high-severity warning

Overview

Best for: Phone and web agents that want Sonic voices with a managed runtime and simple per-minute billing.

At a glance

Flat $/min$0.06
Native S2SNo
ToolsYes
WebRTCNo
WebSocketYes
Phone / SIPYes
Open weightsNo

$0.06/min base excludes LLM tokens; Cartesia phone numbers add $0.014/min. Subscription plans include monthly credits. LLM chosen from a per-account list.

Audio in

pcm_16000, pcm_24000, pcm_44100 (16-bit) or mulaw_8000, headerless mono

Audio out

Streamed audio_output events (speaking-pace delivery option)

Languages

Configurable per agent; list not read.

Voices

Cartesia Sonic voice library.

Latency

Not stated on the pages read.

Regions

Twilio regional routing documented; Cartesia hosting regions not stated.

Compliance

Not stated on the pages read.

Features

  • system, webhook and client tools
  • dynamic variables
  • agent versioning
  • turn-taking and interruptions
  • outbound and batch calling
  • call metrics/evaluations

Pricing

WhatPriceUnitNotes
Agent calling$0.06per minuteBase rate for all voice agent calls; excludes LLM.
Cartesia-provided phone number telephony+$0.014per minute
LLMper model token pricesper 1M tokensListed per account via GET /v1/agents/models. Free LLM usage promotion ran until October 1, 2026.
How the per-minute estimate was worked out

$0.06 base plus $0.014 telephony on Cartesia numbers plus LLM tokens (small models add little; own estimate).

Audio token rate

Not token-billed (LLM portion is).

Free tier: Subscription plans include monthly credits; free LLM promo ended 2026-10-01.

Source: docs.cartesia.ai

Setup

  1. Create an agent in the Cartesia Playground or via API (instructions, voice, LLM, tools).
  2. Server: open the agent WebSocket with X-API-Key and cartesia_version; browsers: mint an agent access token.
  3. Send session_create with input_format, wait for session_ready, stream audio_input, play audio_output, clear on audio_output_clear.

Endpoint

wss://api.cartesia.ai/v1/agents/websocket/{agent_id}?cartesia_version=2026-08-14

Authentication

X-API-Key header (server) or access_token query parameter (browser/mobile)

Quick start javascript

// npm i ws   (create the agent in the Cartesia Playground or via API first)
import WebSocket from "ws";

const ws = new WebSocket(
  `wss://api.cartesia.ai/v1/agents/websocket/${process.env.CARTESIA_AGENT_ID}?cartesia_version=2026-08-14`,
  { headers: { "X-API-Key": process.env.CARTESIA_API_KEY } } // browsers: ?access_token=... instead
);
let ready = false;

ws.on("open", () => {
  ws.send(JSON.stringify({
    type: "session_create",
    audio: { input_format: "pcm_16000", output_delivery: "speaking_pace" },
  }));
});

ws.on("message", (raw) => {
  const ev = JSON.parse(raw.toString());
  if (ev.type === "session_ready") { ready = true; console.log("call", ev.call_id); }
  if (ev.type === "audio_output") play(Buffer.from(ev.audio, "base64"));
  if (ev.type === "audio_output_clear") stopPlayback(); // user interrupted
});

// headerless mono 16-bit PCM, 20-100 ms chunks
export function sendPcm(buf) {
  if (ready) ws.send(JSON.stringify({ type: "audio_input", audio: buf.toString("base64") }));
}
function play(pcm) {} function stopPlayback() {}

Written from the current docs. Check the vendor's SDK version before you ship.

Warnings

Free LLM period just ended

LLM usage was free until October 1, 2026; from now on LLM tokens are billed on top of $0.06/min. Budgets built during the promo are too low.

Versioned API header

Connections require cartesia_version (2026-08-14 in current docs); older version strings may behave differently.

Never ship the API key to clients

Browsers cannot set WebSocket headers; use short-lived agent access tokens.

Cascaded pipeline

STT, LLM and TTS are separate stages; tone and emotion in the caller's voice are not passed to the LLM.

Telephony surcharge

Cartesia numbers add $0.014/min; Twilio import or SIP is billed by your carrier.

Plus 14 warnings that apply to all voice-to-voice APIs. See category warnings.

Limits

  • Concurrency not documented on the pricing page
  • max_output_tokens per response 1-4096

Models and products

NameStatusNotes
Ink-2 (STT) + chosen LLM + Sonic-3.6 (TTS)GACascaded pipeline. LLM list per account via GET /v1/agents/models (e.g. gpt-5.4-mini).

Docs and sources

Docs

Sources used

Not fully verified

Concurrency limits, languages, latency, compliance.

Similar voice-to-voice APIs

Spotted a wrong price or a dead link?