The cap table is the story. On May 5, 2026, ElevenLabs disclosed the third close of its Series D — the same $500M round it first announced in February at an $11 billion valuation — and the new investor list reads like a pension-fund roll call. BlackRock. Wellington. D.E. Shaw. Schroders. Then the strategics: Nvidia, Salesforce Ventures, Deutsche Telekom (via T.Capital), and KPN. Per TechCrunch’s reporting, the round has now stretched past $550M of total capital.

In the same breath, CEO Mati Staniszewski confirmed ARR crossed $500M in Q1 2026, up from roughly $350M at year-end 2025. That’s $100M of net new ARR in a single quarter.

For voice AI as a category, this is the moment the lights go on. The institutional money showing up doesn’t underwrite an interesting demo. It underwrites something it expects to be infrastructure inside three years. If you’re a CIO running an RFP for voice agents, this announcement should change how you weight the shortlist.

Quick Summary: What Happened

Detail	Info
Date announced	February 2026; third close disclosed May 5, 2026
Round size	$500M Series D, $550M+ with extension
Valuation	$11B
ARR	$500M+ in Q1 2026 (up from $350M at year-end 2025)
Net new ARR in Q1	~$100M
Institutional investors	BlackRock, Wellington, D.E. Shaw, Schroders
Strategic investors	Nvidia, Salesforce Ventures, Deutsche Telekom (T.Capital), KPN, Santander
Existing backers	Sequoia, Andreessen Horowitz, ICONIQ
Primary source	ElevenLabs Series D announcement (TechCrunch)

Bottom line: Voice AI just crossed from “interesting pilot” to “the asset managers are buying.” That changes the buyer’s risk profile — and the timeline before consolidation eats the smaller players.

---

What Actually Happened

The round isn’t new. ElevenLabs first announced its Series D in February 2026 at an $11B valuation — Invezz reported the milestone at the time, and the IPO chatter started almost immediately. What landed this week is the third close of that same round, with a meaningfully different investor mix than the first two closes.

Per Music Business Worldwide, the third close brought in the institutional names: BlackRock, Wellington, D.E. Shaw, and Schroders. Those are public-equity asset managers. They generally don’t write checks into private companies that aren’t on a near-term path to a public listing. Their participation is the part of the announcement most worth paying attention to.

The strategic side matters for a different reason. Nvidia gets compute alignment with one of the highest-volume inference customers in voice. Salesforce Ventures gets a foothold in a category that fits neatly into Service Cloud and Agentforce. Deutsche Telekom — through its T.Capital arm — already runs ElevenLabs in production for customer support, in-network live assistance, real-time translation, and marketing video, per the company’s disclosure. KPN sits on the same European telco logic. Santander rounds out the financial-services exposure.

The ARR number tells the operational story. $350M at the end of 2025. $450M by end of Q1 2026. North of $500M when Staniszewski disclosed it. Net new $100M in 90 days, the bulk of it coming from enterprises deploying voice agents in production — not from the consumer audiobook crowd.

The Consumer-to-Enterprise Flip

For most of ElevenLabs’ history, the public face of the company was voice cloning for podcasters and audiobook producers. Look at our own ElevenLabs vs Murf comparison — written when the brand was still mostly known for content creator tooling. That framing is now stale.

The 2026 ElevenLabs is an enterprise voice agent company that happens to also run a consumer business. The revenue mix at the end of 2025 sat around 50/50 between enterprise and consumer. Staniszewski has publicly committed to shifting that to roughly 60/40 enterprise by year-end, and the Q1 numbers suggest it’s happening faster than that target implies.

The product page tells you where the money actually flows. ElevenLabs Agents is a platform for deploying voice and chat agents in support, sales, hiring, and marketing — low-latency conversation in 70+ languages (Eleven v3 / Conversational AI platform), enterprise-grade security, and a deployment model designed to ship in days rather than months. That’s not a creator tool. That’s a contact center replacement story.

If you’ve been tracking the broader enterprise AI deployment market, the playbook is familiar. Frontier capability gets productized into a vertical surface — in this case, voice — and a sales-led organization runs it into the enterprise top of the funnel. ElevenLabs is now operating at scale in that mode, with a sales culture that reportedly demands 20x annual quota from new reps and an enterprise team expanding fast enough to keep pace with the ARR.

Why the Institutional Money Matters

Here’s the part most coverage missed.

BlackRock, Wellington, D.E. Shaw, and Schroders are not venture investors. They are predominantly public-market firms with private allocations sized as pre-IPO crossover positions. When they show up on a private cap table, the read is almost always the same: they’re buying the share they expect to mark up at IPO and then continue holding into the public listing.

That doesn’t mean an ElevenLabs IPO is imminent. The company hasn’t filed, hasn’t announced bankers, and hasn’t telegraphed a timeline. What it does mean is that the underwriters are positioning. An $11B valuation with $500M of ARR is a 22x multiple on revenue — aggressive for any business not growing 100%+ year-over-year, and ElevenLabs is. The crossover funds wouldn’t take the position at that multiple if they weren’t comfortable with the public-market math.

For an enterprise buyer, the institutional money has two practical implications.

First, vendor durability is now meaningfully better. ElevenLabs is no longer a venture-funded company at risk of failing to raise its next round. It’s a crossover-funded company being prepped for the public markets. The single largest vendor risk in voice AI procurement — that your supplier disappears in a downturn — just got smaller.

Second, the platform-consolidation clock is running. The same crossover funds that just bought into ElevenLabs are in OpenAI, Anthropic, Microsoft, and Google. They have a strong financial interest in ElevenLabs not getting eaten by a hyperscaler before the IPO marks. But they have an even stronger interest in voice as a feature inside the bigger AI platforms over a 5-year horizon. That tension is the actual strategic story of the next 18 months in voice AI.

How Buyers Should Read This Before Procurement

If your team is currently evaluating voice AI tools, the May 5 disclosure should reorder your shortlist. Three concrete questions to add to the RFP.

What questions should enterprise buyers ask voice AI vendors after the ElevenLabs Series D?

1. Is the vendor on a credible IPO track or already inside a hyperscaler? Crossover-funded private companies and hyperscaler-owned products are now the safe tier. Pure venture-funded voice startups carry meaningfully more vendor risk than they did six months ago.
2. Does the vendor publish enterprise-grade SLAs for latency and availability? Voice agents have lower tolerance for failure than text agents. A 200ms latency miss in a support call is a customer experience defect; the same miss in a chat agent is invisible.
3. What’s the vendor’s posture on multi-language coverage and regional data residency? Deutsche Telekom and KPN didn’t show up on the ElevenLabs cap table by accident. European telcos need 70+ languages and EU data residency. If your business has cross-border traffic, this is the single biggest functional differentiator.
4. How is the vendor priced for high-volume inference? ARR growth at ElevenLabs’ rate implies aggressive enterprise pricing. The list price you see today is not the price the company will be willing to negotiate against a Nvidia-Blackwell-backed cost structure inside 12 months.
5. What’s the integration story with the rest of the agentic stack? Voice is one modality. The procurement question is whether your voice vendor plays cleanly with your text agent platform, your CRM, and your contact center infrastructure — not whether it sounds the most natural in a demo.

A vendor that can answer all five credibly is selling something different than a vendor that can’t. ElevenLabs can answer most of them today. Most of its competitors can answer two or three.

The Platform Consolidation Risk

The piece nobody at the announcement wants to discuss out loud.

Voice AI as a standalone category has roughly 24 months before the platform players absorb most of the volume. OpenAI ships voice as a native feature of GPT models. Google ships voice through Gemini Live. Microsoft ships voice through Copilot and Teams. The hyperscalers don’t need to buy ElevenLabs to compete in voice — they need to ship “good enough” voice as a free or low-cost feature inside the platform their customers already pay for.

That’s the standard SaaS-feature-vs-platform story, and voice AI isn’t immune to it. The historical pattern: the standalone category leader either gets acquired at the peak, IPOs and reprices to a feature multiple, or pivots to depth in a vertical the platforms can’t service. ElevenLabs is hedging by going deep into specific verticals — conversational AI for support, media and entertainment, music with ElevenMusic, and the developer API.

The Nvidia-Google partnership announced earlier this year fits this hedge. ElevenLabs runs on Nvidia Blackwell GPUs through Google Cloud, which gives it the compute architecture to undercut hyperscaler-native voice on inference cost. Whether that’s enough to keep the category independent through an IPO is the open question. Crossover funds are betting yes. Skeptics are noting that voice is a feature, not a category, and that the same logic applied to standalone search summarization in 2023.

For a buyer in 2026, the practical implication is that you’re choosing between two reasonable bets. Bet on the platform: you get integration depth and pricing power, but second-tier voice quality. Bet on ElevenLabs: you get best-in-class voice and a vendor with serious institutional backing, but you carry the consolidation risk.

We’d argue the latter bet is currently underpriced.

How ElevenLabs Stacks Up on Specifics

Voice quality: Still best-in-class against open-source alternatives like Mistral’s Voxtral and against established players like Murf. The gap has narrowed but hasn’t closed.

Languages: 70+ supported. Deutsche Telekom (Germany) and KPN (Netherlands) are the public reference cases for European production use.

Latency: Sub-second for conversational agents on the Nvidia Blackwell stack — the threshold for natural-feeling phone interactions.

Pricing: Enterprise pricing is undisclosed. Public consumer tiers run $5–$330/month, consistent with the voice generator roundup and voice AI input guide.

Integrations: IBM watsonx Orchestrate, Salesforce, and Twilio are confirmed. The Salesforce Ventures stake suggests deeper Service Cloud and Agentforce integration is coming.

Our Take

The framing for this announcement matters more than the dollar number.

Voice AI is the second category — after coding agents — to reach institutional investability inside the broader generative AI wave. That’s a real distinction. It means the underwriting case has moved from “interesting tech demo” to “infrastructure being adopted at scale by enterprises that pay annually.” The ElevenLabs cap table is the public artifact of that shift.

For buyers, this is the window where vendor selection matters most. Once the category consolidates (it will, on a 24-to-36-month horizon), the procurement power shrinks. A buyer that locks in an enterprise contract with ElevenLabs in 2026 gets best-in-class voice quality, a credibly durable vendor, and pricing power before the hyperscaler-driven race-to-the-bottom kicks in. A buyer that waits until 2028 will be choosing between commodity voice features inside Microsoft 365 and Google Workspace.

We’d go further. The pattern set by the OpenAI and Anthropic PE deployment ventures is going to repeat in voice. ElevenLabs has the cleanest path to building the equivalent forward-deployed engineering team for voice agent rollouts, and the cap table now has the financial firepower to staff it. The next public milestone to watch isn’t an IPO filing. It’s the first $100M+ enterprise deployment contract — likely with a U.S. financial-services firm or a global telco — that ElevenLabs will announce as a case study.

The institutional money is positioned for that milestone. The product is built for it. The only open question is timing.

Frequently Asked Questions

What is ElevenLabs’ valuation in 2026? $11 billion, set in February 2026 at the announcement of the Series D. The third close in May 2026 reaffirmed the valuation at the same level. Per TechCrunch, total capital raised in the round now exceeds $550M.

Who are ElevenLabs’ new institutional investors? BlackRock, Wellington, D.E. Shaw, and Schroders joined the Series D in the third close. Strategic backers added in the same round include Nvidia, Salesforce Ventures, Deutsche Telekom (via T.Capital), KPN, and Santander.

How much ARR does ElevenLabs have? More than $500M as of Q1 2026, up from roughly $350M at year-end 2025. The company added approximately $100M in net new ARR in Q1 2026 alone, primarily from enterprise voice agent deployments.

Is ElevenLabs going public? The company hasn’t filed an S-1 or announced a timeline. The composition of the May 2026 third close — heavy on public-market crossover funds — is consistent with pre-IPO positioning, but no formal listing has been announced.

What does ElevenLabs do for enterprises specifically? ElevenLabs Agents is the enterprise platform: voice and chat agents for customer support, sales, hiring, and marketing operations. Deutsche Telekom uses the platform for customer support, in-network live assistance, real-time translation, and marketing video production. The system supports 70+ languages with sub-second latency on Nvidia Blackwell GPUs through Google Cloud.

Should buyers worry about platform consolidation in voice AI? Yes, but not enough to skip ElevenLabs. The 24-to-36-month consolidation risk is real — hyperscalers will ship native voice features that compete on price. Best-in-class voice quality and the institutional cap table give ElevenLabs durability through that window, and procurement leverage is highest before, not after, the wave hits.

How does this change voice AI vendor selection? The shortlist is now bifurcated. Tier one: hyperscaler-owned voice and crossover-funded specialists like ElevenLabs. Tier two: pure venture-funded voice startups, which carry materially higher vendor risk than six months ago. Weight institutional backing as a procurement variable, not just product quality.

---

Last updated: May 7, 2026. Sources: TechCrunch · Tech.eu · Music Business Worldwide · Invezz · The Next Web · ElevenLabs Enterprise

Related reading: ElevenLabs vs Murf · Voice AI is Faster Than Typing · Mistral Voxtral Open-Source Voice AI · OpenAI and Anthropic PE Ventures · Enterprise AI Deployment 2026