Meta Platforms has quietly become one of Microsoft's largest AI customers, spending hundreds of millions of dollars per year on Azure to consume trillions of AI tokens weekly, according to Bloomberg1,2.
The report, by Bloomberg's Brody Ford and published on August 20, 2026, reveals the scale of Meta's Azure consumption. The spending positions Meta among Microsoft's biggest AI clients, a relationship that has developed without public fanfare.
ANALYSIS The arrangement is notable given that Meta and Microsoft are direct competitors in AI model development and deployment. Meta develops and releases its own Llama family of open-weight models, while Microsoft has its own AI partnerships and products built on top of models from other providers. Meta's decision to route substantial token volume through Azure rather than relying exclusively on its own infrastructure points to a pragmatic calculation about capacity, cost, or capability gaps that its internal systems do not fully address.
The "hundreds of millions per year" figure and "trillions of tokens weekly" consumption rate underscore the sheer computational demand Meta's AI operations generate. Token volume at that scale suggests usage extending well beyond experimentation into production-grade workloads.
For Microsoft, landing a customer of Meta's size and AI sophistication validates Azure's competitiveness as an AI-serving platform, even among hyperscalers that operate their own data centers at massive scale. The revenue contribution from a single customer spending at this level would be material within Azure's AI segment.