Skip to content
Cartesia logo

Cartesia

Real-time speech and transcription models for voice agents, including Sonic TTS and Ink STT with low-latency streaming APIs.

CodingAudio

How it works / How to use

Real-time speech and transcription models for voice agents, including Sonic TTS and Ink STT with low-latency streaming APIs.

  1. Create an account and API key in Cartesia following the official docs.
  2. Pick the model endpoint that matches latency, cost, and capability needs.
  3. Prototype with a short prompt and log token usage.
  4. Add retries, timeouts, and content filters appropriate to your application.

How to prompt

Sonic is Cartesia's low-latency text-to-speech model for live agents.

Stream Sonic TTS for a customer support greeting and follow-up prompts with sub-200ms latency.

Best output tips

  • Set spend alerts where the dashboard allows.
  • Cache stable completions when safe to reduce cost.
  • Rotate API keys on a regular schedule.
  • Verify current plans and limits on the official Cartesia website.

Try this AI

Try Cartesia

Product Details

Pricing, features, limits and latest updates

Cartesia

Real-time speech and transcription models for voice agents, including Sonic TTS and Ink STT with low-latency streaming APIs.

No listed price

Pricing Plans

No listed plans.

Key Features

Voice

voice workflow (cartesia.ai).

API

api workflow (cartesia.ai).

Ideas / Prompt experiences

Share a prompt that worked for you. Username and email are shown with your submission. External links are not allowed.

Example prompt

Streaming TTS response

Prompt

Stream Sonic TTS for a customer support greeting and follow-up prompts with sub-200ms latency.

Short explanation

Sonic is Cartesia's low-latency text-to-speech model for live agents.

Example prompt

Live transcription

Prompt

Pipe microphone audio to Ink STT and return partial transcripts for a voice agent turn-taking loop.

Short explanation

Ink provides streaming speech-to-text optimized for voice agents.

Share your experience

Required fields are marked. Variation, result, and explanation are optional.

Cursor logo

Cursor

Explore

AI coding agent for building software in your editor, terminal, and IDE.

CodingFree / Paid

AutoGPT logo

AutoGPT

Explore

Platform for building and running AI agents with visual Agent Builder, AutoPilot chat, scheduling, and 200+ integration blocks; self-host open source or use hosted Pro/Max plans.

CodingFree / Paid

Hugging Face logo

Models, datasets, and inference for open machine learning.

CodingFree / Paid