ElevenLabs Exits Beta With 28-Language AI Voice Model After $11B Valuation

ElevenLabs has officially exited beta and launched Eleven Multilingual v2, a foundational AI speech model supporting 28 languages. The release comes just weeks after the company closed a $500 million Series D round that valued it at $11 billion.

The new model automatically identifies written text across nearly 30 languages and generates what the company describes as “emotionally rich” speech. More importantly for content creators, voice characteristics—including original accents—remain consistent across all supported languages. A single cloned voice can now produce content in Chinese, Japanese, Arabic, Korean, and 24 other languages without losing its distinctive qualities.

18 Months of Research Pays Off

ElevenLabs spent a year and a half building new mechanisms for understanding context and conveying emotion in synthesized speech. The timing isn’t accidental. With $330 million in annual recurring revenue reported for 2025 and over 1 million users on the platform, the company needed infrastructure that could scale globally.

That scaling got a boost last month when ElevenLabs extended its Google Cloud partnership, gaining access to NVIDIA Blackwell GPUs to handle increased demand.

The language roster now includes Chinese, Korean, Japanese, Turkish, Indonesian, Filipino, Ukrainian, Greek, Czech, Finnish, Romanian, Danish, Bulgarian, Malay, Slovak, Croatian, Classic Arabic, and Tamil. These join the original eight languages: English, Polish, German, Spanish, French, Italian, Hindi, and Portuguese.

Who Benefits Most

Game developers and publishers stand to gain significantly. Indie studios can now localize voice content for international audiences without hiring voice actors for each market. Embark Studios and Paradox Interactive are already using the platform.

Educational institutions get instant audio content generation in target languages—useful for language learning programs and accommodating international students. Audiobook publishers like Storytel, already an ElevenLabs partner, can expand their catalogs into new markets faster.

The accessibility angle matters too. Visual content can now be supplemented with multilingual speech for users with visual impairments, without the cost of recording separate audio tracks for each language.

What’s Next

ElevenLabs plans to introduce a voice-sharing mechanism that lets users monetize their cloned voices when others use them for content creation. CEO Mati Staniszewski framed it as “human-AI collaboration”—though the revenue-sharing details remain unclear.

The company’s trajectory from $100 million valuation in June 2023 to $11 billion in February 2026 reflects broader market appetite for AI voice technology. With $781 million in total funding across five rounds, ElevenLabs has the runway to pursue its stated goal: making content accessible in any language, in any voice.

Image source: Shutterstock