Nvidia is weighing a funding round investment that would value Mercor, a data labeling provider, at $20 billion, according to people familiar with the discussions.

The valuation would double Mercor's worth in under a year—the company was valued at $10 billion in an October Series C round where it raised $350 million. Mercor had commanded a $2 billion valuation in its Series B.

Mercor supplies training data for Nvidia's Nemotron open-source models. The relationship has become material: Mercor earned tens of millions of dollars from Nvidia in the last quarter alone, according to people with knowledge of the arrangement. The size of Nvidia's potential investment and the total funding round size were not disclosed.

Nvidia's interest in Mercor reflects a broader strategy to control key nodes in the AI supply chain. The chipmaker also invested in Scale AI's 2024 funding round, which valued the company at $14 billion, and sources data from Turing and other providers. By backing multiple data suppliers, Nvidia reduces dependence on any single vendor while ensuring access to training data at favorable terms—a calculated bet that data quality will remain a bottleneck in model development.

Mercor, however, faces litigation that could complicate the funding math. In April, the company was served with at least seven class-action lawsuits alleging a data breach exposed contractor information, including job interview recordings, facial biometric data and computer screenshots. Mercor has disputed the claims and told the Wall Street Journal it "strongly dispute[s] the speculative claims in these lawsuits and look forward to presenting the facts at the appropriate time and place."

Nvidia's data investments sit within a larger infrastructure play. In August, the company announced partnerships with six financial institutions to establish independent compute platforms designed to mobilize over $500 billion in third-party capital for AI buildout. Those platforms will offer capital at favorable rates to Nvidia customers deploying large-scale systems—essentially financing the infrastructure Nvidia's chips power.