Cartesia Sonic

Paid

Ultra-low-latency voices for real-time agents.

Real-TimeLow-LatencyAPI

Overview

Cartesia’s Sonic model is built for real-time, conversational voice with extremely low latency and instant cloning. Its state-space architecture targets responsive agents and phone systems. Developers choose it when fast, natural turn-taking matters most.

What makes Cartesia Sonic special?

Focus: Real-time voice generation.
Availability: Commercial / free tier available.
Use cases: Voice agents, phone bots, and live applications.

Visit official site

Related paid models

ElevenLabs

Ultra-realistic voices with emotion and multilingual support.

Murf.ai

Simple and powerful voiceovers in 20+ languages.

Play.ht

Large voice library and exportable TTS.

Back to directory