Labelbox, an AI platform supporting reinforcement learning for frontier models and managed agents for enterprises, has announced more than US$100 million in Series E financing. The round includes both equity and a credit facility and was led by Infinitum Partners, with participation from Wipro Ventures, Andreessen Horowitz (a16z), B Capital, Tau Ventures, Snowpoint Ventures, and others. The company disclosed the funding alongside news that it reached profitability in 2026 while growing more than fivefold year over year.
Funding and Growth
Labelbox works with more than 400 customers, including over 90% of leading U.S. AI labs, hyperscalers, startups, and Fortune 500 companies. The Series E financing will support continued research and engineering across Horizon and Recursion Managed Agents, as well as expansion of the teams bringing these products to customers. The company also plans to deepen its work with leading AI labs and hyperscalers on post-training programs.
Platform Approach
As AI models become capable of performing increasingly complex tasks, the next challenge is enabling them to learn from experience and reliably execute work in real-world environments. Labelbox is addressing this opportunity through two core products, Horizon and Recursion Managed Agents. Horizon focuses on reinforcement learning and evaluation for frontier labs, while Recursion Managed Agents supports enterprise fleets of agents that improve with every run.
Horizon for Frontier Labs
Horizon provides reinforcement learning environments, expert feedback, and evaluation infrastructure that help AI labs train and improve frontier models. Its environments simulate complex, multi-step work across coding, research, and other demanding tasks, enabling models to practice, receive feedback on their performance, and improve through reinforcement learning. Labelbox works with leading AI labs and hyperscalers on post-training programs that rely on these structured practice environments.
Recursion Managed Agents
Recursion Managed Agents enables companies to deploy fleets of AI agents inside the applications their teams already use. Agents can initiate work in response to schedules, alerts, tickets, and other triggers, coordinate across tools, and have their output evaluated against defined standards. Each completed run can inform the agent's memory, helping it improve how it approaches future tasks without manual retraining.
Leadership Perspective
"Every company will soon run on AI systems that learn from their own work," said Manu Sharma, CEO and co-founder of Labelbox. He noted that getting there takes innovation at every layer, from sandbox infrastructure and the agent harness to reinforcement learning post-training. Sharma added that Labelbox brings those layers together in one platform, and this funding lets the company push each one further so that frontier labs and enterprises can build AI that gets better with every run.
Enterprise Adoption
Labelbox already uses more than 50 agents internally across functions including engineering, research, product, sales, marketing, and finance. Earlier this month, the company opened Recursion Managed Agents to enterprise customers, extending its approach to continuous improvement beyond AI research labs. The platform lets enterprises deploy fleets of agents that coordinate across tools, respond to triggers such as schedules and tickets, and improve with every run.
With this Series E funding, Labelbox aims to help both frontier labs and enterprises build AI systems that learn from their own work and improve over time. The company combines agent infrastructure, sandbox environments, and reinforcement learning post-training systems to support the development and deployment of AI systems across demanding real-world tasks. Its profitability, fivefold year-over-year growth, and expanding customer base signal rising demand for AI platforms that enable continuous learning and reliable execution.