Mercor Raises $750 Million at $20 Billion Valuation in New Funding Round
Key Facts
Mercor announced on November 10, 2026, that it had raised $750 million in a Series D round at a $20 billion valuation, according to the company. The round includes a tender offer for employee shares, alongside participation from a group of investors, the company said. Mercor did not disclose how the amount was split between the offer and new capital. The company said the funding would accelerate its investments in infrastructure and research related to model evaluation and post-training. It also plans to build more environments where AI agents can learn complex workflows.
The investor list included Benchmark, Felicis, Fundomo and funds managed by Apollo affiliates, as well as G Squared, General Catalyst, Glade Brook Capital, NVIDIA, Salesforce, Sapphire Ventures, StepStone Group, XTX Ventures and Zoom. Mercor included an employee share tender offer in the round but did not specify its terms or size. The company’s announcement did not disclose the value of each investor’s participation. The funding is the latest in a capital-raising trajectory that has lifted the company’s announced valuation compared with its previous round. In October 2025, Mercor raised $350 million in a Series C round at a $10 billion valuation.
The new valuation is double the one the company announced in October 2025. Mercor said at the time that its Series C valuation was five times its Series B valuation. That places the new round within a series of large increases in the company’s announced valuations, without the current round announcement disclosing further details about its terms. Mercor says the funding will go toward building the infrastructure and research needed to evaluate models and train them after pre-training. It also plans to expand the environments in which agents learn to execute complex workflows.
These plans focus on measuring model performance against enterprise requirements and companies’ actual workflows. Mercor says it works with companies and experts to build tailored evaluations, and that the results help determine how well model performance aligns with client requirements. The results also show where models could benefit from additional training or practice in real-world environments. The company said it spent the previous two years building infrastructure to train advanced models and evaluate them on real-world work using human expertise. It said these capabilities, developed for leading AI labs, have begun to be adopted by major technology companies, developers of open models, application companies and Fortune 500 enterprises.
Mercor says its network includes more than five million experts who are paid to train advanced AI models based on their specialized knowledge. The company began as an AI-powered recruiting platform before shifting to connecting technology companies with experts whose knowledge is used to train, test and improve models and agents, according to Reuters. Its enterprise evaluation business is based on testing models on tasks that simulate each company’s actual work. The company said the United States alone has more than 800 distinct occupations, each with workflows on which models have yet to be evaluated. Mercor sees these workflows as an opportunity to expand testing and training of agent performance.
Mercor also uses models in its own internal operations. It said it spends three times as much on inference as it does on employee salaries, and that the spending pays for itself by improving its team’s performance. Co-CEO Brendan Foody said, "Every job in the economy will need its own AI agents, and every company will need a way to train and evaluate them on its work." Foody links the spread of agents to the need for tools that measure their performance in each organization’s work environment. According to Reuters’ description, the company provides services that connect experts with model training, testing and improvement. The shift comes as Mercor moves from expert services toward offering evaluation and training infrastructure for enterprises.
The round comes amid expanding funding for AI-focused startups. Those companies raised $102 billion, equivalent to 64% of global venture capital, in the third quarter of 2026, according to Crunchbase data cited by Reuters. Jay Das, a partner at Sapphire Ventures, said leading AI labs have so far been the primary buyers of expert-based training data. He added that he expects this to change as open models improve enough for enterprises to adopt them. The forecast points to a potential expansion in the range of organizations using expert data to develop models, Das said.