ElevenLabs Launches v4 Speech Model with Enhanced Expression Control and Support for Over 90 Languages
ElevenLabs has launched its latest speech models, v4 and v4 Turbo, which offer improved expression control, lower latency, and support for over 90 languages. The new architecture enables faster voice cloning—just 10 seconds of audio are needed—and better handling of long text with consistent voice identity. ElevenLabs expanded on the inline tags introduced in its previous version to allow more nuanced expressions through stacked tags. The company has seen significant growth, particularly in enterprise calling services, where over half of its business now comes from large companies. The v4 model is optimized for voice agents and real-time interactions due to reduced latency. ElevenLabs also reported a substantial revenue increase, with an annualized run rate rising from $330 million at the start of 2024 to over $600 million. The startup recently raised $500 million in funding led by Sequoia Capital, valuing it at $11 billion. Rumors suggest a potential follow-up round could push its valuation toward $22 billion.
Tickers: $GOOGL