Hume AI has introduced EVI 2, a new foundational voice-to-voice AI model that promises to enhance human-like interactions. Available in beta, EVI 2 can engage in rapid, fluent conversations with users, interpreting tone and adapting its responses accordingly. The model supports a variety of personalities, accents, and speaking styles and includes multilingual capabilities.
EVI 2 is designed to adapt to user preferences through specialised emotional intelligence training. It aims to provide engaging and personalised interactions by maintaining a consistent voice identity across sessions. Notably, the model avoids voice cloning risks by restricting modifications to its core voice characteristics.
Developers can utilise EVI 2’s experimental voice modulation feature, which allows adjustments to attributes such as pitch and gender without cloning. This innovation supports the creation of customised voices and personalities for specific applications.
The initial release of EVI 2 includes the ‘small’ version. Future updates will enhance its reliability, language support, and instructional complexity. A ‘large’ version is also in development.
You can try out the model here.
Founded in 2021, Hume is a research lab and technology company with a mission to ensure that artificial intelligence is built to serve human goals and emotional well-being. It is founded by Alan Cowen, a former researcher at Google AI.
The company raised a $50 million Series B funding from EQT Group, Union Square Ventures, Nat Friedman, Daniel Gross, Northwell Holdings, Comcast Ventures, LG Technology Ventures, and Metaplanet.
Meanwhile, OpenAI has yet to release its ‘Advanced Voice Mode’ for ChatGPT. This mode enables more natural, real-time conversations that can detect and respond to emotions and non-verbal cues. OpenAI is gradually rolling out access to ChatGPT Advanced Voice to a select group of ChatGPT Plus subscribers and plans to make it available to all paying users by the end of the year.
