ElevenLabs has launched Eleven v3 (alpha), a new version of its text-to-speech model that supports over 70 languages, more than double the 33 languages supported in the previous version. This model offers enhanced expressiveness with features including multi-speaker dialogue with contextual awareness and audio tags such as [excited], [sighs], [laughing], and [whispers]. The company has announced a competition inviting users to showcase the best voice generations created with Eleven v3, with winners receiving Meta Ray-Ban AI Glasses. Concurrently, other companies are advancing voice AI technologies, such as Cartesia AI's Ink-Whisper, a fast and affordable streaming speech-to-text model optimized for real-world conditions, and HeyGen's Avatar IV update, which improves gesture control, facial expressions, and script-driven animation. Additionally, ChatGPT's voice mode received an update in 2025, enhancing human-like conversation with smoother intonation and real-time responsiveness. ElevenLabs also introduced Speech Studio, a tool offering dozens of voices and tones for text-to-speech conversion.