OpenAI's up to $110 billion raise lines up almost exactly with the $111 billion it just added to its cash burn forecast

OpenAI closes the largest private financing round in history. Amazon invests up to $50 billion and becomes a strategic partner. Meanwhile, Microsoft stresses that its own partnership remains unchanged.

OpenAI has announced a $110 billion financing round. That would more than double the company's own record of $40 billion from last year. The pre-money valuation climbs to $730 billion, a major jump from the $500 billion set in a secondary sale in October.

The three main investors are splitting the sum as follows: Amazon is putting in up to $50 billion, while Nvidia and SoftBank are each contributing $30 billion. Additional investors are expected to join as the round progresses, OpenAI said.

A significant chunk of the total hasn't actually been locked down yet, though. Amazon's investment starts with an initial commitment of $15 billion, according to OpenAI. The remaining $35 billion will flow "in the coming months when certain conditions are met." OpenAI didn't specify what those conditions are.

Amazon deal pairs funding with a broad strategic partnership

Alongside the funding, OpenAI and Amazon have announced a multi-year strategic partnership. Under the agreement, the two companies will develop customized models to power Amazon's customer-facing applications.

The existing $38 billion collaboration with Amazon Web Services is being extended by another $100 billion over eight years. AWS will also become the exclusive third-party cloud distribution partner for OpenAI's Frontier enterprise platform.

As part of the financing round, OpenAI is also expanding its collaboration with Nvidia. The company plans to tap three gigawatts of dedicated inference capacity and two gigawatts of training capacity on Nvidia's Vera Rubin systems. These build on Hopper and Blackwell systems already running at Microsoft, OCI, and CoreWeave.

Microsoft wants everyone to know its OpenAI deal isn't changing

The pivot toward Amazon was likely watched closely in Redmond: AWS is Microsoft's biggest competitor in cloud computing. In a joint statement, OpenAI and Microsoft emphasized that the terms of their partnership, which dates back to 2019, will not change "in any way." The partnership "remains strong and central," the companies write.

In practical terms, Microsoft's exclusive license and access to OpenAI's intellectual property stays in place. The revenue-sharing agreement still applies and, according to the statement, also covers revenue from OpenAI's partnerships with other cloud providers.

The key distinction comes down to infrastructure: Azure remains the exclusive cloud provider for stateless OpenAI APIs. Stateless API calls resulting from collaborations with third parties like Amazon are also hosted on Azure. For additional compute and large-scale infrastructure—the Stargate project, for example—OpenAI explicitly has the flexibility to source capacity elsewhere.

OpenAI's own products, including Frontier, will continue to run on Azure. The contractual AGI definition and associated processes also remain unchanged. The fact that both companies felt the need to spell all of this out in a detailed joint statement speaks for itself.

The current investor structure has a notable quirk: both Amazon and Nvidia directly benefit from OpenAI spending the capital it raises right back on them—whether on cloud infrastructure at AWS or GPUs from Nvidia. It's a circular setup where the investors are also the biggest beneficiaries of the investment.

New funding lines up almost exactly with OpenAI's ballooning costs

OpenAI recently revised its cash burn forecast upward again. The company now expects cumulative cash outflows of $665 billion by 2030, roughly $111 billion more than previously estimated. The up to $110 billion announced here lines up almost exactly with that additional funding gap.

OpenAI doesn't expect to hit cash-flow positive until 2030. In 2025, inference costs quadrupled and the adjusted gross margin dropped to 33 percent instead of the targeted 46 percent. Training costs alone are projected to reach nearly $440 billion by 2030.