AI training startup Mercor announced on Tuesday a $750 million Series D funding round that values the company at $20 billion, nearly double the $10 billion valuation it reached just over a year ago. The round was confirmed by Reuters and detailed in the company's announcement distributed via Business Wire, and it includes an employee share tender alongside new institutional capital.
The investor list is a who's-who of the AI supply chain. Nvidia, Salesforce, and Zoom participated, according to the announcement, alongside Benchmark, Felicis, Fundomo, funds managed by affiliates of Apollo, G Squared, General Catalyst, Glade Brook Capital, Sapphire Ventures, StepStone Group, and XTX Ventures. For more context on this story, see our ongoing AI industry coverage.
From Recruiting Marketplace to AI Training Infrastructure
Mercor's rise has been remarkably fast. In September 2024 the company raised a $30 million Series A at a $250 million valuation, according to PR Newswire. By February 2025 it had secured $100 million at a $2 billion valuation, as reported by TechCrunch and CNBC. In October 2025, a $350 million Series C lifted the valuation to $10 billion, with coverage from TechCrunch, CNBC, and the Wall Street Journal.
The company was founded by a group of Thiel Fellows led by CEO Brendan Foody, a Harvard College dropout who started the company as an AI-powered recruiting marketplace before pivoting into the business that defines it today: organizing human experts to train and evaluate frontier AI models. The Wall Street Journal described the founders as 21-year-olds when the company hit its $2 billion valuation, and NDTV later reported that Foody and his co-founders became some of the world's youngest self-made billionaires as the valuation climbed. Business Insider reported in late 2025 that Mercor was paying out more than $1.5 million a day to the humans training AI systems. The company brands itself as "organizing human intelligence to power the AI economy," and the funding announcement makes clear the bet is on depth of expertise, not volume of clicks.
What the New Capital Will Fund
According to the company's announcement, the funding will accelerate infrastructure and research for model evaluations and post-training — the additional training applied to an existing model — while supporting more environments where AI agents can learn complex workflows.
Mercor says it has spent the past two years building infrastructure to train and evaluate frontier models on real-world work using human expertise. It is now extending those capabilities to enterprises, offering custom evaluations that use an employer's actual workflows to measure models against company requirements and identify where additional training or practice in realistic environments could help. The company reports that its network includes more than five million experts paid to train frontier AI models, and that the United States alone has more than 800 distinct occupations containing workflows against which models have yet to be evaluated.
Mercor also opened offices in New York and London in 2026, expanding beyond San Francisco to be closer to enterprise and laboratory customers.
The company's enterprise playbook, as described in the announcement, starts with a capable open model, builds public evaluations, and uses post-training to reduce costs before training models for customers. In other words, Mercor is selling not just labeled data but a measurable pipeline: establish what an agent can do, what it costs, and where targeted training moves those numbers for a specific employer's workflows.
Nvidia's Interest Was Long-Rumored
The Information reported in August 2026 that Nvidia was in discussions to fund Mercor at a $20 billion valuation, and Forbes had earlier reported the company was in talks to raise $500 million at that same valuation. Tuesday's announcement confirms those discussions materialized into a completed, larger round with Nvidia officially among the participants — a notable data point in the ongoing consolidation between chipmakers and the data providers that train their customers' models.
A Company That Has Weathered Scrutiny
The round caps a turbulent year. In April 2026, Mercor confirmed a major data breach, according to Fortune, and Wired reported that Meta paused its work with the company after the incident put AI industry data at risk. Axios reported in September 2025 that Scale AI had sued Mercor for trade secret theft. Even the company's fundraising has not been free of drama: TechCrunch reported in June 2026 that Foody publicly accused Sequoia of "dual-pricing" valuation tricks, an unusual public spat between a founder and one of Silicon Valley's most storied firms. Through all of it, the company continued to grow revenue and investor interest, and Tuesday's valuation suggests buyers of AI training services have not retreated.
A Bellwether for the AI Data Economy
Mercor sits in the middle of one of the AI industry's most competitive niches: the market for expert human data used in post-training and evaluation of frontier models, where it competes with companies like Scale AI, Surge AI, and Turing. Reuters framed the round as the latest test of Wall Street's appetite for AI bets, arriving on the same day Anthropic's trillion-dollar IPO dominated market headlines. If the $20 billion valuation holds up under public-market scrutiny of the sector, expect more capital to flow into the human-data layer that underpins nearly every frontier model release.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
