Nvidia is negotiating an investment of up to $10 billion in the planned IPO of Anthropic, the company behind the Claude models, according to Reuters. Anthropic intends to raise up to $100 billion and reach a valuation of around $2 trillion, which would make the listing the largest IPO in history. For business, the deal shows how the main supplier of AI chips is turning into a financial participant in the market it supplies.
Terms of the placement and the anchor role
Nvidia would enter as an anchor investor, securing shares before the stock begins trading on the open market. The placement is expected to close before the US midterm elections in November. The size of the round is the central figure here: up to $100 billion raised and a valuation near $2 trillion put Anthropic in a category of its own, well beyond the scale of previous technology listings. Nvidia's contribution of up to $10 billion is therefore a small share of the total raise, but it carries weight because of the timing and the status of the buyer.
The relationship between the two companies is not new. Anthropic runs its models on Nvidia GPUs and committed in 2025 to purchasing $30 billion of Azure compute equipped with Nvidia chips. The commercial results behind the listing are visible in revenue: around $9 billion at the end of 2025 and more than $65 billion by July 2026. That trajectory explains both the size of the raise and the willingness of a strategic investor to commit before the shares are priced.
Nvidia's position in the industry goes beyond selling hardware. The company invests in its own customers, including Anthropic and OpenAI, and backs roughly $300 billion in guarantees that help data centers obtain financing. Much of that money returns to Nvidia as orders for chips. The talks with Anthropic fit this pattern: capital goes out to buyers of compute, and demand for GPUs comes back through the same channel.
What this means for business
For companies that build on Claude, the listing changes little in the short term at the level of the product itself, but it changes the financial context. A public Anthropic with a large capital reserve can fund compute purchases and model development without depending on a narrow group of private backers. The practical consequence is predictability of supply and pricing for API customers, especially for those whose processes are tied to a single model provider. A small company notices this mainly through the stability of tariffs and the pace of new model releases; a large one through the terms of long-term contracts and the ability to negotiate volumes.
The risks are concentrated in the same place as the benefits. Nvidia acts simultaneously as supplier, investor and guarantor for its customers, so part of the demand for its chips is financed by its own capital. When assessing such a scheme, a buyer of AI infrastructure should separate two questions: how much the vendor's product costs today and how much of the market's demand rests on money from the vendor itself. The news does not mean that Anthropic's compute or model prices will fall, nor that the placement will take place on the stated terms. Before making decisions, it is worth clarifying with suppliers how contracts are structured, what share of capacity is reserved and for how long.
The marker to watch is the completion of the placement before the November midterms and the final valuation against the announced $2 trillion. If the IPO closes at or near that level with Nvidia among the anchor investors, the scheme in which chip suppliers finance their own buyers becomes the norm for the largest players, and business should expect contracts to be tied to such alliances. If the placement is postponed or the valuation falls short, the model of circular financing will face its first serious test.
