OpenAI raises $110 billion at $730 billion valuation in landmark AI funding round
Massive investments from Amazon, Nvidia, and SoftBank underscore the escalating cost of building global AI infrastructure.
OpenAI said it has secured $110 billion in new investment at a $730 billion pre-money valuation, a big jump from its $500 billion valuation in a secondary round in October.
The funding includes $30 billion each from SoftBank and Nvidia, and $50 billion from Amazon. Additional investors are expected to join as the round progresses. Alongside the financing, OpenAI announced an expanded strategic partnership with Amazon and new infrastructure agreements with Nvidia.
Central to the announcement is a sweeping expansion of OpenAI’s relationship with Amazon, including a multi-year strategic partnership focused on delivering AI tools to enterprises, startups, and consumers worldwide.
Under the agreement, Amazon will become the exclusive third-party cloud distribution provider for OpenAI’s Frontier enterprise platform through Amazon Web Services. Frontier allows organizations to build, deploy, and manage teams of AI agents operating across business systems with built-in governance and security.
The companies are also jointly developing what they describe as a “Stateful Runtime Environment,” a new framework designed to allow AI models to retain context, access memory, and operate across software tools and data sources. These environments will be optimized for AWS infrastructure and integrated with Amazon Bedrock AgentCore services, with launch expected in the coming months.
OpenAI is also committing to consume approximately two gigawatts of Amazon’s Trainium computing capacity as part of a broader expansion of an existing infrastructure agreement that will grow by $100 billion over eight years. The arrangement spans both current Trainium3 chips and future Trainium4 processors, which are expected to begin delivery in 2027 and offer higher performance and expanded memory capacity.
At the same time, OpenAI said it is expanding its longstanding collaboration with Nvidia, securing three gigawatts of dedicated inference capacity and two gigawatts of training capacity on Nvidia’s Vera Rubin systems. These arrangements build on existing deployments using Nvidia’s Hopper and Blackwell systems across cloud providers.
The investment comes amid rapid expansion in OpenAI’s user base and enterprise footprint.
The company said more than 900 million people now use ChatGPT weekly, including over 50 million paying consumer subscribers. Subscriber growth accelerated significantly at the start of the year, with January and February on track to be the strongest months for new subscriptions in the company’s history.
In the workplace, more than nine million paying business users rely on ChatGPT, with organizations deploying the technology across engineering, finance, customer support, and operations.
OpenAI’s coding product, Codex, has also seen rapid adoption. Weekly Codex users have more than tripled since the start of the year, reaching 1.6 million.
