Voice AI startup ElevenLabs completed a $300 million tender offer that let employees sell vested shares at a $22 billion valuation, doubling the $11 billion figure the company reached when it raised $500 million in February.
The transaction was co-led by Wellington Management and T. Rowe Price, two institutional investors that typically hold private company stakes through an eventual public listing. Employees sold a portion of their vested equity directly to those buyers rather than through a new funding round, meaning no fresh capital entered the company's balance sheet.
It is the second employee liquidity event ElevenLabs has run in roughly a year. The New York- and London-based company held a $100 million tender in September 2025 at a $6.6 billion valuation, which means its implied worth has more than tripled in the 12 months since.
Founded in 2022, ElevenLabs builds software that generates synthetic human voices and sound effects. Its tools are used in audiobook production, video game development, advertising, and dubbing, and the company has become one of the most highly valued startups in Europe.
The tender offer reflects a widening practice among fast-growing AI companies that use secondary sales as a retention tool. By letting early employees convert paper equity into cash, startups reduce the incentive for staff to leave for a competitor or a larger tech firm offering liquid stock compensation.
Wellington and T. Rowe Price are both long-established asset managers that invest in late-stage private companies. Their participation at a $22 billion valuation signals institutional appetite for AI voice technology even as questions persist about how synthetic audio will be regulated and how rights holders will be compensated.
ElevenLabs co-founder and CEO Mati Staniszewski spoke with reporters last week but has not publicly detailed revenue figures or a timeline for an initial public offering.
The company's valuation trajectory — from $6.6 billion to $11 billion to $22 billion in about a year — places it among the fastest-appreciating private AI firms outside the foundation model developers. Whether that pace holds will depend on enterprise adoption of voice generation and on how courts and lawmakers treat the training data behind it.
