Open-Source TTS Reaches New Heights: Nari Labs Releases Dia, a 1.6B Parameter Model for Real-Time Voice Cloning and Expressive Speech Synthesis on Consumer Device

Nari Labs has launched Dia, a powerful open-source TTS model using 1.6 billion parameters for real-time voice cloning and expressive speech synthesis. Dia, released under the Apache 2.0 license, supports zero-shot voice cloning, generates non-verbal sounds, and operates efficiently on consumer devices. Its modular design and availability on Hugging Face make it accessible for both commercial and academic use, standing as a strong alternative to proprietary systems.