Nvidia plans to invest in AI data provider Mercor, which is valued at $20 billion.

Nvidia has used Mercor data for training its Nemotron open-source models, paying tens of millions of dollars in service fees last quarter. This investment marks Nvidia's expansion from chip customers to the upstream of the data supply chain. Mercor expects gross revenue to reach $2 billion this year, and after this round of financing, it will surpass Scale to become the highest-valued unicorn in the data annotation field.

Nvidia is seeking to acquire a stake in AI data labeling company Mercor, a move that would double the valuation of the three-year-old startup to $20 billion. This reflects the chip giant's expansion of its investment reach into the upstream of the data supply chain as it accelerates the development of open-source AI models.

According to a recent report by tech media outlet The Information, sources familiar with the matter revealed that Nvidia has begun discussions to participate in Mercor's new round of financing, with existing investor General Catalyst in talks to lead the round. Last quarter, Nvidia paid Mercor tens of millions of dollars for data services, indicating a rapidly deepening business relationship between the two companies.

This potential investment comes against the backdrop of Nvidia's accelerated investment pace. In the quarter ending in April of this year, Nvidia invested a total of $18.6 billion in private companies and infrastructure funds, exceeding its total investment for the entire year of 2025. If Mercor successfully completes this round of financing, it will be Nvidia's first major bet in the data services sector.

Nvidia and Mercor: From Customers to Potential Shareholders

Nvidia's relationship with Mercor is evolving from a simple procurement contract to an equity-based partnership. According to reports, sources familiar with the matter revealed that Nvidia has used Mercor's data to train its two latest Nemotron open-source models, and several Mercor employees are dedicating almost all their time to serving Nvidia.

Mercor's core business involves recruiting human experts from various fields to use and evaluate the performance of AI models in specialized tasks. The resulting data is then used by model developers to improve their models' capabilities. These tasks range from scientific research and legal and financial projects to programming NVIDIA GPUs.

NVIDIA's Nemotron series is central to its open-source model strategy, aiming to compete with the world's top open-source models. As NVIDIA increases its investment in Nemotron, its payments to Mercor also rise, leading to a continued increase in Mercor's revenue contribution to NVIDIA.

It's worth noting that Nvidia is not Mercer's only data provider. The report indicates that, according to sources, Nvidia also uses services from other data providers such as Turing and Scale, and has its own in-house data team.

In addition, NVIDIA also obtains valuable data generated when developers access Nemotron models through the OpenRouter platform. A person who has been involved in the Nemotron project said that, like many AI model developers, NVIDIA relies heavily on synthetic data (i.e., AI-generated data) to train its models.

Mercor: A High-Growth Data-Driven Unicorn

Mercor's financial data confirms the booming AI data annotation market. According to a previous report by The Information, Mercor achieved gross revenue of $614 million in the first half of this year, a 70% increase compared to the entire previous year, and expects its full-year gross revenue to reach $2 billion.

In terms of profitability, Mercor achieved a slight positive operating profit last year and has already forecast to investors that operating profit will rise to $226 million this year and further jump to $1.7 billion next year. Since the company pays contractors approximately 60% to 70% of its gross revenue, its net income for the first half of this year is estimated to be between $180 million and $250 million.

Mercor, founded just three years ago, has already raised approximately $520 million from investors including Benchmark and Felicis, and was valued at $10 billion last October. If this round of financing is completed at a valuation of $20 billion, its valuation will double in less than a year.

Mercor's main customers have historically been developers of closed-source AI models, including OpenAI, Google, and Anthropic. Nvidia's rapid growth indicates that its customer base is expanding towards open-source models.

Nvidia's Investment Landscape: From Chip Customers to Ecosystem Layout

Nvidia's recent investment strategy shows a clear trend of scaling up and accelerating.

This year, Nvidia has invested $2 billion each in cloud service providers CoreWeave and Nebius, both of which are major customers of Nvidia chips. In addition, it has invested in Lumentum and Coherent, suppliers of Nvidia network hardware lasers, and provided funding to several small startups that rent or purchase Nvidia chips.

Nvidia's investment logic typically revolves around its direct customers or the largest users of its chips. The potential investment in Mercor is slightly different—Mercor is not a chip buyer, but rather an upstream data provider for Nvidia's AI model development capabilities. This move reflects Nvidia's strategic considerations in strengthening its open-source AI competitiveness.

For reference, Nvidia participated in Scale AI's 2024 funding round, at which time Scale was valued at $14 billion. Subsequently, Meta acquired a 49% stake in Scale and hired its CEO. If Mercor completes this round of funding, it will surpass Scale with a valuation of $20 billion, becoming the highest-valued company in the AI data annotation field.