Voice AI startup ElevenLabs just got twice as expensive. The company completed a $300 million employee tender offer at a $22 billion valuation on Wednesday, doubling the $11 billion mark it set when it raised $500 million in February. The doubling is being driven less by hype than by workload: ElevenLabs says its AI voice agents now handle more than 15 million conversations a week, triple the February level.

It is a striking number for a company founded in 2022. The New York- and London-based startup, co-founded by Piotr Dabkowski and CEO Mati Staniszewski, started out generating ultra-realistic human voices and sound effects — the kind of tool that powered narrators, dubbing, and audiobooks. Two and a half years on, it is selling something bigger: voice agents that talk to your customers for you.

What a tender offer actually is#

This was not a fundraising round. In a tender offer, investors buy shares from existing shareholders — mostly employees cashing out a portion of their vested equity — without the company necessarily raising new capital. The valuation matters because it resets the price everyone’s options are worth, not because ElevenLabs banked another pile of cash.

The deal was co-led by Wellington and T. Rowe Price, the kind of large institutional investors that typically buy into private companies with the intention of holding on after they go public. Existing backers Andreessen Horowitz and Lightspeed joined in, alongside new names including EQT and Goldman Sachs.

It is the second time the four-year-old company has run a secondary transaction for employees. The first, in September 2025, was a $100 million tender at a $6.6 billion valuation. The step-up between those marks tells the story in miniature:

WhenTransactionValuation
September 2025$100M employee tender$6.6 billion
February 2026$500M Series D raise$11 billion
September 2026$300M employee tender$22 billion
Mati Staniszewski, co-founder and CEO of ElevenLabs
Photo of Mati Staniszewski, co-founder and CEO of ElevenLabs. CC BY-SA 4.0 by Rafał Masłow via Wikimedia Commons.

The numbers behind the doubling#

Valuations doubling in seven months need an explanation, and ElevenLabs’s is usage. Its agents are processing refunds, renewing insurance policies, and booking appointments — the unglamorous, high-volume work that call centers have struggled to staff. The company says its models can speak, listen, and translate in more than 90 languages spoken by over 5.5 billion people, which is why the customer list skews toward enterprises and governments.

“We’re already seeing rapid adoption of expressive voice agents by enterprises and governments, who are deploying them in service of consumers and citizens,” Staniszewski said.

The timing is no coincidence. Reuters notes that some of the biggest AI investments of the year have gone to agent developers, and major tech companies are crowding in — Meta has been expanding its personal AI agent Muse toward small businesses. ElevenLabs’ bet is that whoever owns the voice interface owns a durable wedge into that market, because a phone call is still how a huge share of customer relationships actually happen.

ElevenLabs logo
ElevenLabs logo. CC BY-SA 4.0 via Wikimedia Commons.

The employee-liquidity trend#

ElevenLabs is part of a growing pattern: fast-growing AI startups using secondary transactions as a retention tool. When your private valuation keeps climbing but there is no IPO in sight, vested equity is paper wealth — and talented engineers start taking calls from recruiters. Letting staff sell a slice of it on a schedule keeps them around without forcing an early public listing.

With the new mark, ElevenLabs joins the ranks of Europe’s most valuable startups. That is notable in itself: the voice-AI race has been framed as an American and increasingly Chinese contest, but one of its best-capitalized contenders still builds out of London and New York.

What to watch#

A $22 billion valuation on secondary-market terms is a bet that voice agents keep compounding. The risk is concentration: agent infrastructure is getting crowded, the big model labs keep pushing their own voice capabilities, and enterprise buyers are ruthless about per-minute economics. The thing to watch is whether conversation volume keeps tripling — and whether the margin on those 15 million weekly conversations can survive the pricing pressure that always follows adoption curves like this. If it can, the next headline is an IPO filing. If it can’t, $22 billion is the top of a very expensive bell curve.

Sources