$30 billion a year! That’s what OpenAI is willing to pay just for power and commute. That’s like three times its current revenue. But why? OpenAI has signed the biggest cloud deals in history with Oracle, as reported by TechCrunch. This partnership is much more than just for power and commute. It’s going to mark a turning point in AI infrastructure, cloud competition, and the future of large-scale language models (LLMs).
Details About the Deal
In the month of June, 2025 a Securities and Exchange Commission filing revealed that Oracle had signed a cloud contract, that expected to generate $30 billion in annual revenue. Weeks after that the client remained unnamed, which fueled widespread speculation of who it could be. Then finally in July, OpenAI confirmed the deal through a public blog post from the company and a statement from CEO Sam Altman on X. But the figure amount of the deal hasn’t been confirmed yet.
What We Know About the Deal
- Customer: OpenAI
- Provider: Oracle Cloud
- Estimated value: $30 billion/year
- Scope: 4.5 gigawatts of capacity
- Location: Abilene, Texas (Stargate I site)
To just explain how much 4.5 gigawatts of capacity is let me put it this way- 4.5 GW is equivalent to two Hoover Dams. If I had to break it down into something much simpler, 4.5 GW would be enough to supply energy to 4 million homes!
The Project: Stargate
As reported by The WSJ, this deal supports the Stargate initiative, a $500 billion mega-project launched earlier this year by OpenAI, Oracle, and SoftBank. To just jog your memory let me tell you what the Stargate Initiative is aiming to do. It aims to build the world’s most powerful AI data center infrastructure. That has a combined capacity of 10 gigawatts over time.
And with the help of this partnership, OpenAI will use this infrastructure to power its future LLMs. Which is going to include GPT-5 and beyond, as well as enterprise APIs and tools like ChatGPT Team and ChatGPT Enterprise.
Why This Deal Matters
This deal doesn’t just indicate a win for Oracle, it represents a big strategic leap for OpenAI as well. So it’s like a win-win situation for both of them.
Here’s what makes the deal so significant:
| Benefit | Explanation |
| Scalability | OpenAI will have long-term access to compute power without depending on competitors like Microsoft Azure. |
| Exclusivity | Oracle is becoming a dominant player in the AI cloud race. Thereby, challenging AWS, Azure, and Google Cloud. |
| Acceleration | LLM development, training, and deployment speeds are going improve drastically. |
| Control | OpenAI will gain more autonomy. In factors such as how it handles infrastructure, privacy, and performance tuning. |
Challenges Ahead
Well the road towards achieving success isn’t free from bumps, as there are a few challenges that will need tackling. So let’s dive into the challenges that might occur despite having this much scale:
- Infrastructure cost: Oracle is expected to spend $50 billion in capital expenditures. In how many years, how much time do they have? In just over the next two years to support such growth.
- Energy & environment: Do you know how much energy would be required to power 4.5 GW? Powering 4.5 GW of compute comes with a lot environmental and regulatory hurdles. Check out this article that talks about How Much Energy Does AI Use Amid Rising Environmental Concerns to know more.
- Execution risk: Building and operating such massive data centers is usually time sensitive. To add on to it, it is logistically complex as well.
Nevertheless, OpenAI is known for making big moves, and that’s why now it’s where it is. Despite these challenges and the arising questions of sustainability, one thing is clear: whoever controls compute at scale, controls the future of AI. This contract signals something big: the beginning of a New AI Cloud Era.
