Microsoft relies heavily on OpenAI for its Azure AI revenue, with approximately 70% flowing through this single partnership. In April 2026, Microsoft revised its partnership terms, maintaining primary cloud access while capping the revenue share arrangement through 2030. At its Build 2026 conference, Microsoft unveiled seven proprietary MAI models and emphasized chip co-design to reduce its OpenAI dependence. The company is diversifying through Azure Foundry by partnering with Anthropic, Mistral, and Meta, including investments in European GPU infrastructure. The key metric to watch is whether the 70% OpenAI dependency decreases while total AI revenue grows, which would validate Microsoft’s diversification strategy.
Source: Read the original article
