Key Takeaways

  • ElevenLabs passed a $450 million annual run rate and raised $500 million at an $11 billion valuation after an inflection in voice agent reliability.
  • The company added $100 million in net new ARR in a single quarter, driven by enterprise expansion.
  • Over 50% of ElevenLabs revenue now comes from sales-led enterprise contracts rather than purely self-serve subscriptions.
  • When ElevenLabs releases a flagship model, they subsidize it by selling it at cost to accelerate customer adoption and get immediate performance feedback.

Subsidizing the Flagship Model at Cost

Most software companies price new, high-compute models at a fat margin on day one. They spend months justifying the price tag to enterprise buyers before anyone touches the software. ElevenLabs does the exact opposite.

“The way we usually do is when we have a new model, we try to give it at cost to a lot of the customers so they can experience the best,” Staniszewski explained to John Collison on Cheeky Pint.

Selling at cost sounds like a fast way to burn cash. In practice, it solves the hardest problem in AI product development: latency and quality discovery. When developers get access to frontier voice tech without a pricing penalty, they build actual products instead of small weekend experiments. That usage gives ElevenLabs immediate, high-volume telemetry on edge cases, latency bottlenecks, and real conversational failures. By the time competing providers catch up to the technical quality, developers have already wired ElevenLabs into production workflows.

The Shift to Sales-Led Enterprise Revenue

ElevenLabs started as a self-serve platform for creators and developers. But pure developer self-serve rarely sustains nine-figure annual growth on its own. The real revenue acceleration happened when the team built a dedicated sales motion on top of their developer base.

“We are over 50% is now sales-led enterprise,” Staniszewski noted. That shift paid off immediately: “This quarter was one of the best for enterprise growth where we had the first quarter hit 100 million in additional ARR growth, which is crazy.” Staniszewski pointed out that while the company previously announced reaching a $350 million run rate at the end of 2025, that rapid enterprise expansion quickly pushed them past $450 million. They backed up that momentum with a $500 million fundraise at an $11 billion valuation.

The self-serve product is no longer the entire business. Instead, self-serve acts as an unpaid qualification filter for the sales team. Developers test the voice models, build functional prototypes inside their companies, and hit usage ceilings. When an enterprise contract arrives, the engineering team has already vetted the API. The sales team just needs to solve security, procurement, and volume pricing.

The Dual-Engine Trap to Avoid

Running a self-serve funnel and an enterprise sales team at the same time kills most early-stage startups. Founders try to build custom enterprise features for five-figure deals, which starves the core product team of resources.

ElevenLabs made this work because the underlying product is an API. The core model that serves an individual hobbyist is the same model that powers an enterprise call center. Enterprise buyers are not paying for bespoke features; they are paying for reliability, throughput, and dedicated capacity. If your product requires two completely different codebases to serve individual developers and enterprise accounts, this hybrid model will split your company in half.

What to Do With This

Take your newest or highest-performing feature and run a 30-day cohort where you offer it to your top 20% most active users at raw cost. Require one thing in exchange: an automated log of latency and failure points sent directly to your engineering team. Use the surge in usage to fix edge cases and build pipeline for enterprise contracts next quarter.