ElevenLabs launches v4 speech models, now in more than 90 languages

ElevenLabs released two new speech models on Monday, ElevenLabs v4 and v4 Turbo, adding more expression control, lower latency for voice agents, and support for over 90 languages. The company released its v3 model last year and teased the new generation at an event in Warsaw earlier this year. For v4, ElevenLabs adopted a new architecture that allows for better control and faster cloning — with v4, the company says users will be able to clone a voice with just 10 seconds of audio.

On the creative side, the model handles voice identity better over longer chunks of text and keeps the context of the text in mind while reading it aloud, adjusting expression accordingly. ElevenLabs introduced inline tags to define expression with v3; v4 expands those tags, letting users stack multiple tags and have the model follow the sequence. The previous version supported 70 languages; v4 pushes that to more than 90, with the biggest quality gains in Japanese, Brazilian Portuguese, Mandarin and Cantonese.

ElevenLabs has scaled its enterprise calling business rapidly over the past year, with more than 55% of its business now coming from large companies, and the company says v4 is built for voice agents: the new version has lower latency for more fluid conversation, and can start generating audio as soon as the underlying LLM starts generating its answer. The model can also handle confrontations, escalations and holds differently to improve issue resolution. Competition in speech models has intensified, with startups including Cartesia, Deepgram, Fish Audio, Boson and WellSaid Labs building expressive speech models of their own, while Google and OpenAI have also improved their voice models.

ElevenLabs raised $500 million earlier this year in a round led by Sequoia that valued the company at $11 billion; there are already rumors of a follow-up round that would value it at $22 billion. The company’s annualized revenue run rate has climbed from roughly $330 million at the start of the year to over $600 million, and its headcount has grown past 800 as it has hired aggressively in markets including India, Europe and Brazil. In a recent interview with TechCrunch, co-founder and CEO Mati Staniszewski said the company is aiming for an IPO “in the next years,” but didn’t commit to a timeline.

Why it matters

ElevenLabs is betting that voice-agent latency and language breadth — not just voice realism — decide who wins enterprise calling contracts, with its biggest v4 quality gains landing squarely in Brazilian Portuguese and other non-English languages Technopolis’s own readers use.