# Mercor Doubles Valuation to $20B With $750M Series D as It Takes AI Agent Evaluation Beyond the Frontier Labs
> Mercor raises a $750M Series D at a $20B valuation, double its last round, to train and test AI agents on enterprise work with Nvidia and Salesforce.
- **Author:** Prashant Jagtap
- **Published:** 2026-11-11
- **Updated:** 2026-11-11
- **Category:** Startups
- **Editorial type:** news
- **Topics:** series-d, nvidia, salesforce, ai-startups
- **Canonical URL:** https://www.todaysstartupnews.com/startups/mercor-raises-750m-series-d-20b-valuation-enterprise-ai-agent-evaluation
## Key takeaways
- Mercor has raised $750 million in a Series D round at a $20 billion valuation, double the $10 billion it reached after a $350 million Series C in October 2025
- No lead investor was named, and the round includes an employee tender whose size was not disclosed
- Nvidia, Salesforce and Zoom took part, alongside returning backers Benchmark, Felicis and General Catalyst
- Mercor plans to sell custom evaluations and post-training environments to enterprises, building on its network of more than five million experts
- The company says its annualised revenue passed $2 billion in June and that it now has about 700 employees
Mercor, the San Francisco company that pays experts to train and test AI models, has raised $750 million in a Series D round at a $20 billion valuation. The company announced the round on November 10, 2026. The valuation is double the $10 billion Mercor reached when it raised a $350 million Series C led by Felicis in October 2025.

Mercor did not name a lead investor. The round drew corporate money from Nvidia, Salesforce and Zoom, and returning investors Benchmark, Felicis and General Catalyst also took part. Other participants were Fundomo, G Squared, Glade Brook Capital, Sapphire Ventures, StepStone Group, XTX Ventures and funds managed by affiliates of Apollo. According to a blog post by co-chief executive Brendan Foody, the round included a tender offer for employee shares, and Mercor has not said how the $750 million splits between that and new capital.

Mercor was founded in 2023 by Foody, Adarsh Hiremath and Surya Midha, and started as an AI hiring platform. Foody and Hiremath are co-chief executives. The company now connects technology firms with subject-matter experts whose knowledge is used to train, test and improve AI models and agents. Its network includes more than five million experts. The company says it has about 700 employees, and Foody said in July that its annualised revenue had passed $2 billion in June.

So far, most of Mercor's work has gone to frontier AI labs. The new funding is aimed at enterprises in every industry, which Mercor argues need their own way to measure whether an AI agent can be trusted with a job. Foody has said every job in the economy will need its own AI agents, and that companies will need a way to train and evaluate those agents on their own work. In a post announcing the round, Mercor said that over the next five years more money will be spent applying AI to real work than on building new models.

Hiremath said a model can score well on a public benchmark and still fail on the systems, policies and edge cases that define how a specific business runs. For each enterprise customer, Mercor builds a custom benchmark that measures models against that company's own standards, and post-training starts from that benchmark. The approach builds on APEX, Mercor's public benchmark series that measures AI performance on economically valuable professional tasks. According to the company, the United States has more than 800 distinct occupations whose workflows models have not yet been evaluated against.

Part of the post-training work runs on software from Deeptune, the New York startup Mercor acquired in July. Deeptune's team had spent the two years before the acquisition recreating hundreds of enterprise applications that AI agents can practise on. Mercor said the new money will pay for research and evaluation infrastructure, including more of these training environments. The company also says it spends three times as much on inference, the computing used to run models, as on employee salaries.

Jai Das, president, partner and co-founder of Sapphire Ventures, said frontier labs have been the main buyers of expert-driven training data, but that this should change as open models become good enough for enterprises to build on. He said Mercor has built something hard to replicate through several years of work with the labs.

Mercor competes in a growing market for specialised AI training data, human feedback and evaluation. Scale AI is a major rival, having raised $1 billion in a Series F at a valuation of nearly $14 billion in 2024.