According to ElevenLabs (@elevenlabsio), the new Eleven v3 (alpha) model is now the most expressive AI Text to Speech solution, introducing multi-speaker dialogue with advanced contextual awareness. This update expands language support from 33 to over 70, significantly increasing global accessibility for businesses deploying AI voice solutions. Additionally, v3 supports audio tags such as [excited], [sighs], [laughing], and [whispers], enabling more nuanced and natural voice synthesis for industries like entertainment, education, and customer service seeking to leverage hyper-realistic AI voices for multilingual, context-rich audio applications (Source: ElevenLabs Twitter, June 12, 2025).