Labelbox has raised more than $100 million in Series E financing to expand the infrastructure that lets AI models and enterprise agents learn from their own work. Why it matters: the money is flowing to the unglamorous layer — training environments and managed agents — that decides whether AI actually improves after deployment.
October 15, 2026 — San Francisco. Labelbox announced the round on Wednesday. It combines equity and a credit facility, was led by Infinitum Partners, and drew participation from Wipro Ventures, Andreessen Horowitz (a16z), B Capital, Tau Ventures, and Snowpoint Ventures, among others.
What the money is for
The funding backs two products. Horizon provides reinforcement-learning environments, expert feedback, and evaluation infrastructure that frontier AI labs use to train and improve their models — simulated, multi-step work across areas like coding and research, where a model can practice, get scored, and improve. Recursion Managed Agents, opened to enterprise customers earlier this month, lets companies deploy fleets of agents inside the business applications they already use. Those agents can start work from schedules, alerts, or tickets, coordinate across tools, and feed each run back into their memory so later runs get better.
Labelbox is not a small customer of its own pitch: the company says it already runs more than 50 agents internally across engineering, research, product, sales, marketing, and finance.
The business behind it
Labelbox says it serves more than 400 customers, including over 90% of leading U.S. AI labs, reached profitability in 2026, and grew more than fivefold year over year. Those are the company’s own figures from its announcement, not audited numbers, but the investor list — and the size of the round — suggests the market believes the post-training layer is where durable value sits as raw model capability commoditizes.
Why it matters
Most AI tools you subscribe to are frozen at the version you bought. Labelbox’s bet is that the next competitive edge is systems that improve with every run — and that both labs and ordinary enterprises will pay for the plumbing that makes that safe and measurable. For readers of this site, the practical signal is where enterprise AI spending is heading: less toward one-off chatbots, more toward agents embedded in existing workflows with evaluation attached. If your business is evaluating agent platforms, “how does it learn from its own output, and how is that checked?” is quickly becoming the question that separates real products from demos.
Frequently asked questions
How much did Labelbox raise, and who led it? More than $100 million in Series E financing (equity plus a credit facility), led by Infinitum Partners.
What does Labelbox actually sell? Reinforcement-learning environments and evaluation tools for AI labs (Horizon), and managed AI agents that enterprises deploy inside their existing applications (Recursion Managed Agents).
Is Labelbox profitable? The company says it reached profitability in 2026 while growing more than fivefold year over year; those figures come from its own announcement.
Sources: Labelbox announcement via PR Newswire (October 14, 2026); SaaS News funding summary.