Over the past few years, we’ve watched text-based AI transform entire industries. Meanwhile, the most common mode of human connection—voice—hasn’t gone through the same evolution. Even in a tech world filled with workhorse chatbots, purpose-built conversational agents, and steady improvements in underlying model capabilities, voice-based AI has been hampered by multiple limitations. The biggest of them? The limits of synthetic speech quality.
Now, that’s finally changing. And ElevenLabs is among those leading the charge. They ended 2025 at $300M+ in ARR and have one of the strongest financial profiles we have seen for a business of this scale.
We’re thrilled to announce our investment in ElevenLabs’ Series D.
Voice AI is a Mega-Trend
Humans spend a third of their waking hours in some form of speech communication—far more time than we spend reading and writing combined. Being able to transform that communication into synthetic speech taps into a significant addressable market.
What makes this moment different: B2B synchronous voice agents have reached production quality. We’ve seen the evidence across our portfolio and the broader market: voice-first companies are achieving meaningful inflection, enterprises are transforming their operations, and the ROI on labor displacement is proving out in real deployments.
Revolut deployed ElevenLabs Agents to enhance customer support across the UK and Europe, scaling coverage to more than 4 million of its 70 million global customers. The results: 8x faster time to resolution, 99.7% call success rate, and native-quality voice support in 31+ languages, available 24/7.
The implications of voice AI extend far beyond customer support. Internal training, sales, marketing, advertising—any function that relies on human speech communication is now in scope for automation and augmentation.
A Voice AI Company Building Up the Stack
Founded in 2022 by Mati Staniszewski and Piotr Dabkowski, ElevenLabs began by developing a truly human-like AI text-to-speech model. Since then, the company has expanded beyond voice, advancing novel research across speech-to-text, sound effects, dubbing, music, and conversation. Today, ElevenLabs brings these models together with integrations, orchestration, and enterprise-grade infrastructure to deliver production-ready platforms for businesses, creators, and developers worldwide.
Their vision is ambitious but increasingly validated: voice is quickly becoming the dominant medium for human-computer interaction as synchronous agents become increasingly viable alternatives to human labor. Model improvement is TAM-expansionary: as voice quality improves, more use cases are unlocked.
Multiple Acts: From Model Company to Platform
What excites us most about ElevenLabs is their evolution from a pure voice model company to a multi-product platform aggressively moving “up the stack”. Each layer of ElevenLabs’ stack gives the layer above it a greater right to win. This is a journey we at Lightspeed have been closely following at our portfolio company Anthropic.
Eleven Creative is a platform for creators and brands to generate, edit, and localize high-fidelity audio across more than seventy languages. Brands like Duolingo, NVIDIA, and the Financial Times use Eleven Creative to create commercial voiceovers and music and localize content for global audiences.
Eleven API provides low-latency, production-grade voice infrastructure for developers building interactive products and experiences. Companies including Meta, Epic Games, Salesforce, MasterClass, and Harvey use Eleven API to power platforms reaching over one billion aggregate users.
Eleven Agents enable enterprises to deploy voice and chat agents with the reliability, integrations, testing, and monitoring required for large-scale customer operations. It’s an open platform for developers and enterprises to create custom-tailored conversational agents. Organizations including Deutsche Telekom, Square, the Ukrainian Government, and Revolut use Eleven Agents for customer support, citizen engagement, internal training, and inbound sales.
The Right Team for This Moment
Mati and Piotr started out as high school best friends in Poland. We’ve watched them both excel when it comes to ambition, learning speed, and a unique perspective on their industry. And they’ve built a team that hails from Google, Meta, and other top companies, bringing both technical depth and enterprise experience to a market that’s evolving fast. Together, they bring an exceptional vision for what the product experiences of the future will look like and how human-technology interaction altogether will change as a result of voice AI technology.
As a testament to all of this, in just four years, they’ve built a category-defining voice AI company, set the bar high for talent dense teams, and consistently executed faster than the market expected.
Why We Invested
Voice AI is a mega-trend that we believe will unlock massive enterprise value in a net-new market. ElevenLabs has built a clear category leader—with powerful models, demonstrated multi-product execution, an exceptional financial profile, and a founding team that has the ambition and attributes to build a generational company.
Voice will be to communication what LLMs are to intelligence—and ElevenLabs is positioned to win that future. When exceptional teams attack massive markets with differentiated technology at the right moment, good things happen. We’re proud to partner with Mati, Piotr, and the entire ElevenLabs team as they build voice for the AI era.
Authors
