Mercor raised a $750 million Series D at a $20 billion valuation, doubling its worth in a year as it shifts from AI hiring into training and testing AI models.
Key Takeaways
- Mercor raised $750M at a $20B valuation on November 10, 2026, up from $10B in October 2025.
- No lead investor was named. Benchmark, Felicis, General Catalyst, NVIDIA and Salesforce took part.
- Funds go to evaluation and post-training infrastructure, plus environments where AI agents learn enterprise workflows.
Lead
Mercor, the San Francisco company that began as an AI-driven recruiting marketplace, announced a $750 million Series D on November 10, 2026. The round values the startup at $20 billion, twice the $10 billion set by its $350 million Series C in October 2025. Part of the round was an employee tender. Mercor did not publish a lead investor.
Who Invested in the Round?
The participant list mixes venture funds, chip and software strategics, and asset managers. It includes Benchmark, Felicis, General Catalyst, Sapphire Ventures, G Squared, Glade Brook Capital, XTX Ventures and Fundomo. Funds managed by affiliates of Apollo and StepStone Group also joined.
On the corporate side, NVIDIA, Salesforce and Zoom invested. Those three are customers or neighbors of the AI model supply chain, which gives the cap table a commercial flavor beyond pure financial backing. The lack of a named lead is unusual for a round this size. It suggests demand exceeded the allocation and that terms were set with little need for a single anchor.
What Does Mercor Actually Do Now?
Mercor now sells human expertise to AI developers. It pays a network of more than five million vetted experts, including doctors, lawyers, engineers and bankers, to produce training data, grade model outputs and build test scenarios for frontier AI labs, technology companies, AI startups and Fortune 500 businesses.
The company was founded in 2023 by Brendan Foody, Adarsh Hiremath and Surya Midha, who serve as co-CEOs. It has about 700 employees, according to TechFundingNews. Only a fraction of the vetted pool works at any time, with roughly 30,000 active contractors in rotation.
The recruiting origin still matters. Mercor's screening and matching tools are how it finds and ranks experts, which is the core of its pitch to model builders. The hiring product is now infrastructure for a different business.
Why Is Mercor Worth $20 Billion?
Revenue growth is the main argument. TechFundingNews reported an annualized revenue run rate above $2 billion as of July. Contractor payouts reportedly absorb 60 to 70 percent of that, which would leave net revenue of roughly $600 million to $800 million. At $20 billion, the valuation sits at about 25 to 33 times that net figure.
That multiple is steep for a business whose largest cost is people, and whose customers are a handful of AI labs with the power to shift volume. Revenue concentration is the obvious risk. A lab that brings data labeling in-house, or moves to synthetic data, could cut a large share of demand quickly.
The doubling in valuation within 13 months also resets the last round. The Series C investors have paid for that confidence with a paper gain of 100 percent. It also leaves the Series D buyers needing another doubling to earn a venture-style return, or an IPO at a higher mark.
How Will Mercor Spend the Money?
Mercor says the capital will accelerate investment in evaluation and post-training infrastructure and research. It will also build more environments where AI agents learn complex enterprise workflows. Earlier in 2026 it acquired Deeptune, a company that builds simulations of enterprise applications for agent training, and it opened offices in New York and London.
The company also said it spends three times more on inference than on employee salaries. That figure points to heavy internal use of AI models, and it shows that compute is a larger expense line than payroll.
Where Does Mercor Sit Among Rivals?
The data-for-AI field is crowded and richly valued. Scale AI drew a reported $14.3 billion investment from Meta for a 49 percent stake. Surge AI has reportedly weighed its first outside raise at a valuation above $25 billion. Snorkel AI is valued at $3.5 billion, and micro1 at $500 million after a $35 million raise in September 2025.
Mercor's differentiator is its talent graph, a database of vetted professionals built before the training-data boom. Rivals can copy the workflow. Rebuilding a network of five million screened experts takes longer.
Outlook
Mercor has doubled its valuation in 13 months and turned a hiring marketplace into a supplier of expert judgment for AI labs. The round is large, the investor list is broad, and the revenue run rate is substantial, but the business depends on a small set of buyers and a high share of revenue paid out to contractors. The next test is whether evaluation and agent-training environments create recurring enterprise revenue beyond lab contracts, and whether synthetic data erodes the human-labeled market that built the company.



