Voice & music
Cartesia
Real-time, expressive text-to-speech API built for voice agents.
Product insights
Cartesia at a glance
Value proposition
Low-latency streaming voices that sound human in conversation.
Problem solved
Slow or robotic TTS breaks the feel of a live voice conversation.
Audience
Developers and companies building voice products.
Pricing and access
Usage-based API pricing with a free tier; see the official pricing page.
Market
- Voice AI
- Developer API
Tech stack
- Sonic TTS model
- Streaming API
01Why it is here
What is Cartesia?
Cartesia AI voice generator is a text-to-speech API. Its Sonic model streams audio fast enough for live conversations, which is why it is used inside customer-support bots, AI companions, and other voice agents.
02Workflow
How to use Cartesia
- 01
Get an API key
Sign up and create a key in the dashboard.
- 02
Pick a voice
Choose a voice and language in the playground.
- 03
Stream in your app
Call the streaming API from your agent.
03Capabilities
Key features of Cartesia
- Sonic-3.6 streaming TTS
- 44 languages
- Expressive output including laughter
- Built for real-time agents
04Use cases
What is Cartesia best for?
- Voice customer support
- AI companions and NPCs
- Narration with low latency
05Questions
FAQ about Cartesia
What is Sonic?
Sonic is Cartesia’s text-to-speech model; Sonic-3.6 is the current version.
How many languages does it support?
Cartesia lists 44 languages for Sonic.
Official website snapshot
cartesia.ai
Cartesia provides Sonic, a streaming text-to-speech model now at Sonic-3.6. It generates natural, expressive voices, including laughter, in 44 languages with low latency, and is designed for AI agents and interactive apps.
