Mercor raises $350M Series C at $10B valuation, paying out $1.5M/day to marketplace experts

Summary

Mercor has closed a $350 million Series C at a $10 billion valuation, a milestone the company frames as validation of its position at the center of AI training infrastructure. The marketplace is now paying out over $1.5 million per day to domain experts, a figure that signals both scale and the intensity of demand from frontier model developers.

The company's strategic framework is openly modeled on Amazon's flywheel logic: more candidates, better matching, faster delivery. The analogy is deliberate. Just as Amazon optimized around product selection, price, and speed, Mercor is building around the same structural constants for AI training marketplaces, where the variables that matter most are expert supply, match quality, and throughput.

The core thesis is that knowledge work is undergoing a structural shift. Rather than performing repetitive tasks directly, workers across industries are increasingly being paid to create evaluations that train agents to handle those tasks autonomously. Mercor positions its marketplace as the infrastructure layer enabling that transition, spanning software engineering, finance, law, and medicine.

Near-term revenue concentration mirrors NVIDIA's early hyperscaler dependence, with roughly a dozen large AI labs driving the bulk of current volume. The longer-term bet is enterprise fragmentation: every large company will eventually need custom agent training for its specific workflows, and Mercor aims to supply the expert labor and eval infrastructure to make that possible.

The quality control argument is central to the competitive moat. Reliable reinforcement learning environments require consistent, expert-reviewed data, a constraint that makes crowdsourcing from end users impractical and creates structural demand for Mercor's vetted specialist network. Disintermediation risk from platforms like OpenAI recruiting experts directly is acknowledged but dismissed on quality grounds.

On market size, Mercor pegs the addressable opportunity at a share of the $40 trillion per year currently spent globally on knowledge work, describing current penetration as roughly 1% of the transformation ahead. An IPO is described as "potentially on the horizon," stopping short of a firm denial without committing to a timeline.