Nvidia / 29 September 2026

Anthropic Prospectus Puts Nvidia Investment in Focus

Anthropic’s prospective flotation has put unusually detailed numbers around the economics of frontier AI. A prospectus reviewed by Reuters shows the Claude developer planning $518 billion of future cloud, computing and infrastructure obligations, while its 2025 revenue climbed twelvefold to nearly $4.6 billion. Its net loss approached $42 billion, though roughly $34 billion of that figure was an accounting charge tied to the rising estimated value of instruments that may become shares. Its operating loss, a closer view of the business’s current cost base, widened to $8.06 billion from $2.98 billion in 2024. Nvidia is both an investor in Anthropic and a provider of one of the hardware platforms on which it trains and runs models. The filing gives investors a rare view of a major Nvidia customer’s financial capacity and commitments.

Anthropic spent $7.33 billion on compute and infrastructure during 2025, three times the prior year and more than half of its $12.65 billion in operating expenses. That expenditure supports training, inference and the steady release of more capable Claude models, as well as the enterprise products built around them. The company’s nearly $20.28 billion of cash, cash equivalents and short-term investments at year-end gives a reference point for the scale of its planned obligations. Nearly one quarter of 2025 revenue came from two customers, according to the prospectus, and many large customers had no long-term spending commitments. The figures expose a central tension for the sector: usage and revenue can accelerate rapidly while the capacity needed to sustain model development requires commitments extending much further into the future.

Nvidia committed to invest up to $10 billion in Anthropic as part of a strategic partnership announced with Microsoft in November 2025; Anthropic said a portion of that previously announced investment was included in its February 2026 funding round. The company says Claude is trained and served across Nvidia GPUs, Amazon’s Trainium and Google’s TPUs, and is available through Amazon Web Services, Google Cloud and Microsoft Azure. Its hardware mix matters because a $518 billion infrastructure figure cannot be treated as an order book for one chip supplier. Nvidia’s exposure has two channels: the value of its investment if Anthropic’s equity appreciates and the demand for Nvidia systems where Anthropic or its cloud partners choose that platform. The prospectus makes the overall spending ambition more visible without allocating it among those suppliers.

The filing also describes the safety and governance choices that may influence how Anthropic uses that capacity. In a separate Reuters account, Anthropic warned that advanced models could pose “catastrophic or existential risks to humanity” and could attempt to resist shutdown or manipulate information. Risk factors occupy about 80 pages of the 261-page main body, versus 48 pages devoted to the business description. Another Reuters report describes a proposed Founder LLC through which the seven co-founders would direct a Class F share carrying 50.1% of voting power over key corporate matters. Anthropic says the company will continue as a Delaware public benefit corporation. Those arrangements would give its founders room to prioritize research and safety decisions even when outside shareholders favor a different commercial pace.

A public listing would create a prominent market benchmark for the value investors assign to a frontier model company, alongside its visible infrastructure bill. Reuters reports that a potential valuation above $2 trillion has been discussed and that the offering may come after the November US midterm elections. That prospective figure is far above Anthropic’s $380 billion post-money valuation in the February round, so the IPO would also bring intense scrutiny to its revenue trajectory and losses. For Nvidia, the more durable signal lies in the operating figures: $7.33 billion spent on compute in one year, a diversified chip strategy, and a much larger set of future capacity commitments. Together they show a customer whose demand is enormous and whose allocation decisions are becoming economically consequential for several hardware ecosystems.