Mercor is a two-year-old startup that focuses on providing domain experts needed by companies like OpenAI and Meta to help train and optimize their foundational AI models. According to a market document obtained by TechCrunch and sources, Mercor is currently discussing its third round of financing (Series C) with investors.

The company's current goal is to increase its valuation to $1 billion or more, which is an increase from the $800 million valuation discussed a few months ago. Although the final terms of the deal may still change, Mercor has informed potential investors that it has received multiple offers, some of which have valuations as high as $1 billion. In addition, Mercor has introduced at least two new investors through special purpose vehicles (SPVs) to raise funds for potential deals.

GPU chip (7)

Image source note: The image was generated by AI, and the image licensing service is Midjourney

Mercor's previous round of financing, announced in February, was a $100 million Series B round with a valuation of $200 million. According to sources, Mercor's annual recurring revenue (ARR) is close to $450 million. Earlier in February, Mercor had publicly announced that its annual revenue reached $75 million, and its CEO Brendan Foody stated on social media in March that ARR had reached $100 million.

The company said it expects to reach a $500 million annual revenue milestone faster than another startup Anysphere. Anysphere is known for its AI coding assistant Cursor and achieved $500 million in annual revenue about a year after launching its product. Unlike Anysphere, which is still burning money, Mercor made a profit of $6 million in the first half of the year.

Mercor's revenue comes from providing professional domain experts to enterprises to perform AI model training, such as scientists, doctors, and lawyers, charged by the hour. The company claims to be a data annotation contractor for five major AI labs, including Amazon, Google, Meta, Microsoft, and OpenAI. According to sources, part of Mercor's revenue comes from OpenAI among these brands.

To further diversify its business model, Mercor is telling investors that it plans to add more software infrastructure for reinforcement learning. This training method allows models' decisions to be verified or refuted, thereby incorporating feedback and continuously improving. The company also intends to eventually establish an AI-driven hiring marketplace.

Despite competition from companies like Surge AI, which is also negotiating to raise funds at a valuation of $25 billion, Mercor remains optimistic. Co-founder Foody said, "We are not trying to raise funds," and rejects multiple investment offers every month.

The three co-founders of Mercor are all Harvard dropouts in their early twenties. Recently, they also hired Sundeep Jain, former Chief Product Officer of Uber, as the company's first President to help the company develop.

However, Mercor is also facing a lawsuit from competitor Scale AI, which accuses it of stealing trade secrets, claiming that a former employee stole over 100 confidential documents related to Scale's customer strategies and other proprietary information before joining Mercor.