On February 4, 2026, voice AI company ElevenLabs announced a $500 million Series D at an $11 billion valuation, led by Sequoia Capital. A year earlier, in January 2025, the company had raised $180 million at a $3.3 billion valuation — so the price of the business more than tripled in twelve months. Since its founding in 2022, total funding now stands at $781 million across five rounds.
The number that matters most here is not the raise itself but the repositioning behind it: ElevenLabs is rewriting its story from “voice tool for content creation” to “enterprise voice agent platform.” TechCrunch and CNBC both report the company closed 2025 with more than $330 million in annualized revenue, and co-founder Mati Staniszewski says the company is building “toward IPO and beyond.”
A Round That Triples the Valuation
Beyond the Sequoia lead — partner Andrew Reed joins the board — the most telling signal is how hard existing investors doubled down: Andreessen Horowitz quadrupled its position, and ICONIQ tripled down. New investors include Lightspeed Venture Partners, Evantic Capital, and BOND, while existing backers BroadLight, NFDG, Valor Capital, AMP Coalition, and Smash Capital also participated. CNBC’s coverage also notes that NVIDIA is among the company’s supporters.
The math behind the pricing is straightforward: a valuation moving from $3.3 billion to $11 billion in a year, resting on $330 million-plus of ARR and an enterprise adoption curve, not on model demos alone.
Three Platforms: Agents, Creative, API
The commercial structure now runs on three layers. ElevenAgents targets enterprise voice and conversational agents — Deutsche Telekom, Square, Revolut, and the Ukrainian government are customers — across customer support, commerce, training, and sales. ElevenCreative handles audio generation, editing, and localization in more than 70 languages, with Duolingo, NVIDIA, and TIME on the roster. ElevenAPI is the low-latency developer infrastructure layer, used by Meta, Epic Games, Salesforce, MasterClass, and Harvey; the company says platforms built on it reach over one billion users combined.
The core model line spans text to speech, speech to text, sound effects, dubbing, music, and conversational AI — both ends of the audio pipeline, input and output, are productized.
Use of Funds: Conversational Models and Global Expansion
The $500 million has three main destinations. First, expanding ElevenAgents, including a major voice agent upgrade: faster responses, improved expressiveness, new turn-taking behavior, and the Eleven v3 Conversational model. Second, research — emotional conversational models, dubbing, and what the company calls audio general intelligence. Third, international go-to-market, with the team spreading across 14 cities including London, New York, Tokyo, São Paulo, and Singapore.
Voice AI’s Enterprise Turn
Place this round in industry context and the pattern is clear: voice is shifting from “the last mile of content production” to “the primary interface for agents.” When an agent needs to make a phone call, answer support lines, or hold a live conversation, voice stops being a feature and becomes the product. That changes the pricing logic too — per-minute API billing is giving way to per-agent, per-seat enterprise contracts, and switching costs on those contracts run far higher than an API price comparison.
The $11 billion valuation is a bet on exactly that tool-to-platform conversion. Sequoia and the insiders who quadrupled and tripled down are wagering that enterprise voice agents become a standard budget line — the way contact-center software did — within the next few years.
What It Means for Developers and Product Teams
Three practical implications. First, if voice is central to your product, revisit your procurement logic: a platform contract (agent orchestration, turn-taking, call interruption handling) locks you in at a different depth than a raw TTS API, so price the exit before you enter. Second, low-latency conversation is the next competitive axis — benchmark Eleven v3 Conversational’s turn-taking and expressiveness in your own scenarios rather than on vendor demos. Third, open-source speech models are maturing at the same time; for privacy-sensitive or cost-sensitive workloads, a hybrid of self-hosted and API speech infrastructure is becoming the norm. Design your vendor abstraction well — it is more reliable than betting on a single platform.
Sources
- Series D announcement — ElevenLabs
- ElevenLabs raises $500M from Sequoia at an $11 billion valuation — TechCrunch
- Nvidia-backed AI voice startup ElevenLabs hits $11 billion valuation — CNBC
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.