ElevenLabs' new v4 speech model makes AI voices more expressive and consistent

ElevenLabs has launched its new speech model, Eleven v4, which enhances voice expression by accurately tracking cues for laughter and whispering. The new model features a Turbo variant that begins speaking in 150 milliseconds, making it suitable for real-time voice agents. Additionally, Eleven v4 ranks first on Artificial Analysis' Voice Arena leaderboard, surpassing both Cartesia and Google's Gemini. This model represents a significant advancement in improving voice quality for long productions like audiobooks, enhancing user experience and enabling new applications in the AI field.