Mercor, a startup that connects companies like Openai and Meta with domain experts needed to train and refine basic AI models, will discuss with Series C round investors, according to two sources that are familiar with transaction consultations and marketing documents seen by TechCrunch.
The company is currently targeting valuations of over $10 billion, one said. This has increased from the $8 billion target valuation the company discussed a few months ago, one said. However, the terms of the final transaction are subject to change.
Mercor’s previous round was announced in February. This is a $100 million Series B, valued at $2 billion, led by Felicis. Felicis declined to comment on the new round.
The company has told potential investors that they already have multiple offers. According to previously reported information, VCS has reached out to Melkor preemptively.
TechCrunch also understands that at least two new investors are trying to raise funds for potential transactions through special purpose vehicles (SPVs). However, we understand that there are no SPVs yet that have been officially approved by the company.
Founded in 2022, Melkor is approaching $450 million in annual occupancy revenue, one said. The company told TechCrunch in February that its annual revenue (calculated by multiplying the latest month by 12) reached $75 million at the time. In March, Melkor CEO Brendan Hoody appeared on X, with an ARR of $100 million.
The company told investors that hitting $500 million milestones faster than AnySphere, a startup that creates AI coding assistant cursors, is on track, according to a source familiar with the situation. Anysphere reached an ARR of $500 million, about a year after the product was launched. Unlike Anysphere, which is still burning cash, Mercor generated $6 million in profits early in the year, Forbes reported.
TechCrunch Events
San Francisco
|
October 27th-29th, 2025
Mercor offers businesses specialist domain experts to carry out AI model training, such as scientists, doctors, lawyers, and other professionals, earning revenue by charging hourly discoverers’ fees and matching rates for their work.
The company claims to provide data labeling contractors to five top AI labs, including Amazon, Google, Meta, Microsoft, Openai and Nvidia. According to sources, a large portion of its revenue comes from a subset of these brands, including Openai.
To further diversify its business model, Mercor is telling investors that it is adding a software infrastructure for reinforcement learning. This is a training method in which the model or agent decision is validated or contested, and incorporates feedback so that it can be improved over time. The company will also ultimately build an AI-powered recruitment market.
Still, Mercor faces competition from companies like Surge AI. This faces competition from Turing Labs and other data labeling companies, including Scale AI, which is expanding to RL services, during discussions to raise funds at a $25 billion valuation. Some believe Openai’s recently launched recruitment platform allows the AI giant to create RL training services with its own human-powered people.
When reaching for the comment, Foody told TechCrunch “We’re not trying to raise them at all,” and “we turned down the offer every month.” He also said the company’s ARR is over $450 million. However, he revealed that the company’s revenue includes the total amount that customers pay to Mellocor for services before the contractor receives the portion. He added that this is a common accounting practice recommended by auditing companies and used by competitors in Surge AI and Scale AI.
The startup was co-founded in 2023 by Thiel Fellows and Harvard Dropouts Brendan Foody (CEO), Adarsh Hiremath (CTO), and Surya Midha (COO). All three co-founders are in their early 20s. To take the company to the next level, Mellcole recently appointed Sundeep Jain, Uber’s former chief product officer with decades of experience, as well as the first president, Forbes reported.
Melkor was recently sued by a competitor’s Scale AI for misappropriation of trade secrets. Scale AI claims that one former employee who later joined Mercor “stolen over 100 confidential documents about Scale’s customer strategy and other unique information,” according to a previously reviewed copy of TechCrunch.
Maxwell Zeff provided the report.
Fix: Previous versions of the story incorrectly described the progress of potential SPVs in this transaction. That information has been updated. The story also revised the list of Melkor’s major customers.
