AI news

Mercor announces $750 million Series D, reaching a $20 billion valuation

By 孙小雯1 viewsSource: 智东西
Mercor founders and illustrations accompanying its financing announcement

According to the company and investors, the round includes employee share sales. The funding will support model evaluation, post-training infrastructure and work environments for agents.

Article illustration

The “data factory” behind OpenAI and Anthropic.

Zhidongxi reported on November 11 that US AI training-data unicorn Mercor announced the previous evening that it had raised $750 million (approximately RMB 5.038 billion) in Series D financing, with NVIDIA, Zoom, Benchmark and others participating. Following the round, its valuation reached $20 billion (approximately RMB 134.334 billion), double the $10 billion valuation (approximately RMB 67.167 billion) following its Series C in October last year. The transaction reportedly also includes arrangements for employees to sell shares.

Mercor said the funding would go toward model evaluation, post-training infrastructure and related research, helping customers test whether models can perform work and improve their performance through further training. The company also plans to build more environments in which agents can practice complex workflows, extending the training and evaluation services it previously provided to frontier AI laboratories to more enterprises.

In the financing blog published at the same time, Mercor cofounder and co-CEO Brendan Foody disclosed that the company’s inference spending to run AI models had reached three times its employee payroll, but the gains from AI improving the team’s work performance were already sufficient to cover that investment.

Founded in 2023 and headquartered in San Francisco, Mercor connects experts across different fields to provide technology companies with training data and model evaluation services. It uses professionals’ knowledge to train, test and improve models and agents. According to Mercor, its expert network covers more than five million people, and all five of the world’s leading AI laboratories, including OpenAI and Anthropic, are its customers.

These experts provide models with experience and judgment from specific kinds of work. According to Mercor, doctors can help AI recognize early warning signs that might otherwise be missed, banking professionals can train agents to perform financial analysis, and lawyers can help models improve their reasoning about case law.

Mercor was founded in 2023 by three people born in the 2000s—CEO Brendan Foody, CTO Adarsh Hiremath and COO Surya Midha—who all dropped out of college to start the company together. They were only 19 years old when they founded it. The three became friends on their high-school debate team. Hiremath had attended Harvard University, while Foody and Midha had attended Georgetown University.

Article illustration

▲ CTO Hiremath, CEO Foody and COO Midha, from left to right

Before this round, Mercor had raised approximately $484 million in total (about RMB 3.248 billion). In October 2025, the company completed a $350 million Series C (approximately RMB 2.351 billion), valuing it at $10 billion (approximately RMB 67.167 billion). In February that year, it completed a $100 million Series B (approximately RMB 672 million); both rounds were led by Felicis. In September 2024, Mercor said it had completed a $30 million Series A (approximately RMB 202 million) led by Benchmark. In January that year, it disclosed $3.6 million in financing (approximately RMB 24.18 million) led by General Catalyst.

On November 10, Sapphire Ventures disclosed in its announcement of this investment that Mercor’s annualized gross revenue had exceeded $2 billion (approximately RMB 13.433 billion), and that it was generating positive free cash flow.

According to Sapphire, Mercor initially used AI to help companies recruit talent before shifting to providing expert training data for AI models. The company still uses AI-led interviews to screen experts.

Mercor said it is currently building customized evaluations around enterprises’ actual work, judging whether agents can follow internal rules and complete specific tasks, then using the findings to conduct post-training and practice in simulated environments. Foody believes that every industry will need AI agents capable of performing specific jobs, and that companies will also need to train and evaluate them using their own business tasks.

Conclusion: AI startups attracted nearly RMB 700 billion in the third quarter

Data from fields such as law and finance remains relatively scarce

According to Crunchbase, AI startups raised a total of $102 billion (approximately RMB 685.103 billion) in the third quarter of 2026, accounting for 64% of global venture capital investment. During this financing boom, supporting services such as training data and model evaluation also attracted capital. Mercor is targeting enterprise applications of agents, providing professional-task training data and evaluations for models. Companies expect AI to master industry knowledge, use internal systems and complete work according to business rules, creating demand for specialized data services.

Mercor and its investors believe that public web data does not adequately cover professional experience in fields such as law and finance, and also lacks enterprises’ internal operating procedures. Professionals can use that expertise to design tasks, create data and evaluate model performance. Ultimately, though, companies purchasing these services still need to see how well AI can do the work. For Mercor, training and evaluation must not only help customers improve task quality and reduce manual rework, but also consider how AI’s benefits can cover the investment required.

Sources: Mercor, The Information

This article comes from “Zhidongxi” and was written by “Sun Xiaowen”.

Related topics

After reading

Join the AINRK community

Read how to join the community and connect with people working in AI.

Community information (opens in a new tab)