Labelbox Raises $100M+ at $1.25B Valuation for AI Agents
Labelbox has raised more than $100 million in Series E financing at a $1.25 billion pre-money valuation, as the AI infrastructure company expands its products for model training, evaluation and enterprise automation. The round combines equity financing with a credit facility and was led by Infinitum Partners, with participation from Wipro Ventures, Andreessen Horowitz, B Capital, Tau Ventures, Snowpoint Ventures and other investors.
Announced on October 14, 2026, the financing will support research and engineering across Labelbox's Horizon and Recursion Managed Agents products, alongside expansion of the teams responsible for bringing them to customers.
The company says it reached profitability in 2026 while growing more than fivefold year over year. It also reports serving more than 400 customers, including over 90% of leading US AI labs.
Horizon Supports AI Training and Evaluation
Labelbox's Horizon platform provides reinforcement learning environments and evaluation infrastructure for AI developers. It allows models to practise complex tasks in simulated environments, receive feedback on their performance and improve through repeated training.
The environments can replicate multistep work involving coding, research, computer use and other business activities. Evaluation systems assess whether a model completes a task correctly, while expert feedback can help distinguish between different valid outcomes.
This approach is intended to help AI laboratories assess and improve models on tasks that more closely resemble real-world work, rather than relying exclusively on static training examples or conventional benchmarks.
Recursion Brings AI Agents Into Enterprise Workflows
The funding will also support Recursion Managed Agents, which Labelbox opened to enterprise customers earlier in October. The platform enables businesses to deploy AI agents inside existing applications and assign them tasks triggered by schedules, alerts, tickets and other events.
Agents can coordinate work across business tools, while separate evaluation systems assess their output against predefined standards. The results can then inform future runs, helping organisations refine how agents complete recurring tasks.
Labelbox says more than 50 agents already operate internally across functions including engineering, research, product, sales, marketing and finance. The company aims to extend this approach to enterprise customers seeking to automate complex workflows while maintaining measurable performance standards.
Investment Targets the Infrastructure Behind AI
Labelbox is investing across several layers of AI development, including agent infrastructure, isolated computing environments, reinforcement learning systems and expert feedback networks. The company argues that reliable AI systems require more than capable foundation models: they also need ways to practise tasks, evaluate results and improve from experience.
Access to computing capacity is expected to remain important as the company expands these products. The combination of equity and credit financing gives Labelbox additional resources for development and commercial growth, although the company has not disclosed how much of the round comes from each financing component.
Competition in AI Infrastructure Intensifies
Labelbox operates in a market where AI companies are investing in training data, model evaluation and agent deployment. Its strategy connects model-development infrastructure with tools designed to help businesses use AI systems in everyday operations.
The latest round gives the company additional funding to expand both areas. Its ability to sustain growth will depend on customer adoption, the reliability of its products and whether enterprises can demonstrate measurable benefits from deploying AI agents.
