According to Reuters, Nvidia is in talks to become one of the anchor investors in Anthropic’s planned IPO. The terms have not been finalised and may still change. Anthropic declined to comment, while Nvidia did not immediately respond to a request for comment.
If the offering reaches the reported size, Anthropic would set a new fundraising record for an IPO. Saudi Aramco, which currently holds the record, raised about $29.4 billion in its 2019 listing after exercising an over-allotment option. The amount reportedly being considered by Anthropic would be more than three times larger.
The AI company is reportedly seeking to complete the IPO before the US midterm elections in November. It confidentially filed paperwork with the US Securities and Exchange Commission (SEC) in June and has since selected Nasdaq for the potential listing. A confidential filing does not guarantee that the transaction will proceed on the reported timetable or terms.
Nvidia’s participation could help Anthropic establish strong initial demand for what may become the largest IPO in history by proceeds. In return, Nvidia would deepen its relationship with one of the AI industry’s major buyers of computing infrastructure.
A $2 trillion valuation would represent another extraordinary jump even by Anthropic’s rapid growth standards. At the end of May, the company completed a $65 billion Series H led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital, bringing its post-money valuation to $965 billion. Three months earlier, its valuation had reportedly been around $380 billion.
Revenue has been rising at a similarly rapid pace. Anthropic’s annualised revenue reportedly increased from approximately $9 billion at the end of 2025 to more than $47 billion in May and over $65 billion by the end of July 2026. Second-quarter revenue was approximately $11.5 billion, about 14 times the year-earlier level. The company reportedly projects revenue of $190–200 billion in 2028, with Claude Code and demand from enterprise customers expected to remain the main growth drivers.
Behind those figures is an intensifying race for computing capacity. Anthropic has entered into a series of agreements with Amazon, Google, Broadcom, Microsoft and infrastructure providers to secure access to chips and data centres.
In August, Riot Platforms announced a 20-year agreement to provide 191 MW of data-centre capacity at its Rockdale, Texas campus to an unnamed leading frontier AI laboratory. Bloomberg later identified the customer as Anthropic. Riot’s official presentation values the initial term at approximately $9.1 billion and the potential total at $16.1 billion if both five-year extension options are exercised.
Anthropic is also pursuing a multi-chip strategy. Its capacity agreements include next-generation Google TPUs developed with Broadcom, Amazon Trainium accelerators and access to Nvidia GPUs, reducing reliance on any single computing platform while supporting further model development and deployment.
The potential Anthropic investment would extend Nvidia’s expansion strategy. In early September, Nvidia agreed to acquire Hugging Face for $12.93 billion, subject to customary closing conditions. The platform serves more than 18 million developers and hosts over 3 million AI models, making it one of the industry’s most important repositories for open models and development tools.
As the AI race accelerates, warnings about its risks are becoming more urgent. Anthropic CEO Dario Amodei recently called on frontier AI companies to slow the development of increasingly capable models and put safety ahead of speed. OpenAI CEO Sam Altman and xAI founder Elon Musk publicly supported the proposal. At the same time, technology competition between the United States and China remains intense, with Washington increasingly focused on export controls and the alleged use of model distillation by Chinese laboratories. These developments add regulatory and geopolitical uncertainty to the otherwise ambitious outlook surrounding Anthropic’s potential IPO.
