OpenAI and Microsoft have agreed to cap total revenue-sharing payments at $38 billion, according to The Information via Reuters. The cap replaces what was previously an uncapped arrangement and gives OpenAI permission to distribute its products across any cloud provider, including Amazon Web Services and Google Cloud.
The Deal Structure
OpenAI will continue paying Microsoft 20% of its revenue through 2030, according to CNBC, citing a source familiar with the agreement. That rate stays the same. What changes is the ceiling: once cumulative payments reach $38 billion, the obligation ends.
Microsoft confirmed the cap’s existence in an April 27 blog post without disclosing the dollar figure. The post stated that the partnership has been “simplified,” that Microsoft remains OpenAI’s primary cloud partner, and that its license to OpenAI’s intellectual property continues through 2032 on a non-exclusive basis.
Two notable clauses were removed. Microsoft no longer receives an exclusive IP license. And the AGI determination clause, which would have allowed OpenAI to stop payments if it achieved artificial general intelligence, is gone. Payments now continue “independent of OpenAI’s technology progress,” per the companies’ joint statement.
Why Now
The renegotiation follows months of visible strain. In an internal memo reported by CNBC earlier in April, OpenAI revenue chief Denise Dresser wrote that the Microsoft partnership had “limited our ability to meet enterprises where they are.”
OpenAI has been building alternatives. In February, Amazon committed up to $50 billion in investment, and OpenAI expanded its existing $38 billion AWS agreement by $100 billion over eight years. AWS became the exclusive third-party cloud distribution provider for OpenAI’s enterprise platform Frontier. Amazon CEO Andy Jassy confirmed on X that OpenAI models would be available through AWS Bedrock within weeks.
Microsoft’s total investment in OpenAI exceeds $13 billion since 2019. In the October 2025 recapitalization, Microsoft valued its stake at $135 billion, roughly 27% on a diluted basis, and OpenAI committed to spending $250 billion on Azure cloud services.
The Cloud Distribution Shift
The most consequential change for agent builders: OpenAI products can now ship on any cloud provider. Azure remains the primary partner, and OpenAI products will ship first on Azure unless Microsoft decides otherwise. But the exclusivity lock is broken.
For enterprises running agent infrastructure across multiple clouds, this removes a friction point. OpenAI’s models were previously available on Azure and through direct API, but enterprise distribution deals were tethered to Microsoft’s ecosystem. That constraint is now gone.
Microsoft, meanwhile, gets financial predictability. A 20% revenue share on a company growing as fast as OpenAI could become an enormous liability if uncapped. At $38 billion, Microsoft knows its maximum exposure and can model returns against its $13 billion investment with precision.
The Infrastructure Market After Exclusivity
The deal closes the era of exclusive frontier model distribution. OpenAI on AWS. OpenAI on Google Cloud. OpenAI on Azure. Enterprise buyers now choose infrastructure based on latency, pricing, and existing contracts rather than which cloud happens to have a lock on the model they need. For teams building agent orchestration layers that span providers, the $38 billion cap is less about OpenAI’s finances and more about what it unlocks downstream: model access without cloud lock-in.