ElevenLabs Launches v4 Speech Models with 90-Language Support and $11B Valuation

ElevenLabs has launched its v4 and v4 Turbo speech models, introducing support for 90 languages, improved expression control, and lower latency for voice agents. The company, which recently raised $500 million at an $11 billion valuation, is doubling down on enterprise applications and global expansion.

ElevenLabs unveils v4 and v4 Turbo speech models

ElevenLabs has launched its latest speech models, v4 and v4 Turbo, which introduce support for 90 languages, up from 70 in the previous version. According to TechCrunch, the models aim to improve expression control, voice cloning, and latency for real-time applications like voice agents.

The v4 models adopt a new architecture that enhances voice identity consistency, even over longer stretches of generated speech. Users can clone a voice with as little as 10 seconds of audio, a significant improvement for applications requiring rapid deployment of synthetic voices.

Inline tags for defining expression, introduced in v3, have been expanded in v4. These tags allow users to fine-tune emotional delivery, making synthetic speech sound more natural and contextually appropriate.

Improved language quality and enterprise focus

The v4 models show notable quality improvements in languages like Japanese, Brazilian Portuguese, Mandarin, and Cantonese. This expansion aligns with ElevenLabs’ push into global markets, where demand for localized speech synthesis is growing.

Over 55% of ElevenLabs’ revenue now comes from large enterprise customers, reflecting its focus on scalable, high-quality solutions for businesses. The lower latency in v4 Turbo is designed to enable more fluid conversations, allowing the model to generate audio as soon as the underlying language model begins producing text.

Funding and growth

ElevenLabs raised $500 million earlier this year in a round led by Sequoia, valuing the company at $11 billion. The funding reflects strong investor confidence in its technology and market potential.

The company’s annualized revenue run rate has nearly doubled this year, climbing from $330 million to over $600 million. ElevenLabs has also scaled its team to over 800 employees, with aggressive hiring in markets like India, Europe, and Brazil.

According to TechCrunch, ElevenLabs’ CEO, Mati Staniszewski, has indicated plans for an IPO in the coming years, though no specific timeline has been shared. Rumors also suggest the company may pursue an additional fundraising round, potentially doubling its valuation to $22 billion.

Background

ElevenLabs, founded in Warsaw, specializes in AI-driven speech synthesis, competing with companies like Google, OpenAI, and Deepgram. Its technology is used in applications ranging from audiobooks to real-time voice assistants.

The company’s previous model, v3, was released last year and supported 70 languages. The v4 models build on this foundation, addressing limitations in voice consistency and multilingual quality.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

ElevenLabs’ latest product launch underscores its ambition to dominate the AI speech synthesis space. The v4 models’ support for 90 languages and enhanced expression control could solidify its position as a go-to solution for enterprises, especially in global markets like Japan, Brazil, and India. The reduced latency for voice agents also suggests a push into real-time applications, such as customer service or interactive AI assistants.

The $11 billion valuation and rapid revenue growth reflect investor confidence, but the company’s aggressive hiring and expansion plans will test its ability to scale without compromising quality. For founders, ElevenLabs’ trajectory highlights the potential of specialized AI models—particularly those addressing multilingual and enterprise needs—but also the challenges of maintaining margins in a competitive space.

Key takeaways

• ElevenLabs’ v4 and v4 Turbo models support 90 languages, up from 70 in the v3 model.
• The new models improve voice cloning, expression control, and latency for real-time applications.
• ElevenLabs raised $500 million at an $11 billion valuation, with revenue growing from $330 million to over $600 million this year.
• Over 55% of ElevenLabs’ revenue comes from large enterprise customers.
• The company is expanding globally, with over 800 employees across markets like India, Europe, and Brazil.

FAQ

What are the key improvements in ElevenLabs’ v4 models?

The v4 models support 90 languages, up from 70 in v3, and offer better expression control, lower latency, and improved voice cloning with as little as 10 seconds of audio. The architecture also ensures more consistent voice identity over longer segments of speech.

How does ElevenLabs’ funding impact its growth?

The $500 million funding round, led by Sequoia, values ElevenLabs at $11 billion and supports its expansion into new markets. The company has grown its team to over 800 employees and doubled its annualized revenue run rate to over $600 million this year.

What industries could benefit from ElevenLabs’ v4 models?

The v4 models are particularly useful for enterprises in customer service, audiobook production, and real-time voice assistants. The support for 90 languages and improved multilingual quality also make them attractive for global applications.

Related on Lazyfounder

• Google
• OpenAI
• More technology news