Coqui Releases XTTS: A Trailblazing Foundation Model for Generative Voice AI

Coqui, an AI startup founded by industry visionaries from Mozilla, has unveiled their groundbreaking foundation model for generative voice AI, called XTTS. XTTS builds on Coqui’s previous success in code and model releases, offering even broader impacts for users. The model, now fully integrated with TTS, supports speech in 13 different languages and leverages advanced generative AI techniques for faster and more refined speech.

The XTTS model introduces several groundbreaking features. It allows voice cloning from just a 3-second audio clip and enables emotion and style transfer during the cloning process. It also has cross-language voice cloning capabilities and supports multi-lingual speech generation. Additionally, XTTS offers a superior 24kHz sampling rate, providing unparalleled voice generation quality.

Hugging Face, a renowned AI platform, has partnered with Coqui to host this transformative model. The collaboration aims to make XTTS universally accessible to developers, researchers, and creative entities worldwide. Coqui, in partnership with open-source licensing expert Heather Meeker, has also developed the Coqui Public Model License (CPML) to ensure transparency, collaboration, and open AI advancement.

Overall, the release of the XTTS model marks a significant milestone in voice AI technology. It bridges the gap between open-access and proprietary systems, setting a new standard for globally accessible voice AI. Coqui invites global innovators to explore and redefine the frontiers of generative voice AI with XTTS, fostering a world where revolutionary technology is universally accessible and driving humanity’s collective evolution.

Sources:

– Coqui AI Press Release: https://www.newsfilecorp.com/release/182483