MODELS
Sonic 3.5
by Cartesia
Overview
Sonic 3.5 is Cartesia’s fast, natural streaming TTS model with sub-90ms latency, 42-language support, and stable dated snapshots.
Details
Sonic 3.5 is a Cartesia text-to-speech model described in Cartesia’s documentation and website as a fast, natural streaming TTS model. The supplied sources state that it supports sub-90ms latency and 42 languages, and that stable dated snapshots are available for production voice agents. The documentation URL is specifically for Cartesia’s latest TTS models page and is titled “Sonic 3.5.”
When to Use
Use Sonic 3.5 when you need a Cartesia streaming text-to-speech model described as fast and natural. Use it for voice-agent use cases where the supplied Cartesia source’s stated sub-90ms latency and stable dated snapshots are relevant. Consider it when multilingual TTS coverage is needed based on the source-stated 42-language support.
Getting Started
- Open the Cartesia Sonic 3.5 documentation page: https://docs.cartesia.ai/build-with-cartesia/tts-models/latest
- Review the model description
- latency claim
- language support
- and snapshot information in the official Cartesia docs.
- Use the Cartesia website for provider-level context: https://cartesia.ai/
- Verify pricing
- usage limits
- and production requirements directly with Cartesia before deployment.
Key Features
- •Streaming text-to-speech model
- •Described by Cartesia as fast and natural
- •Sub-90ms latency stated in supplied Cartesia source metadata
- •42-language support stated in supplied Cartesia source metadata
- •Stable dated snapshots for production voice agents stated in supplied Cartesia source metadata
Capabilities
- •text-to-speech
- •streaming TTS
- •multilingual speech generation
Last updated Jun 19, 2026