Microsoft closed fiscal 2026 with Azure surpassing $100 billion in annual revenue for the first time.
Microsoft announced on July 29 that Q4 revenue reached $90.0 billion, up 18% year over year, while operating income reached $40.6 billion, also up 18%. Net income was $35.8 billion, up 31% on a GAAP basis, and diluted earnings per share was $4.81, up 32% on a GAAP basis.
For the full fiscal year, Microsoft revenue reached $331.8 billion, up 18%, while operating income reached $155.2 billion, up 21%. Microsoft Cloud revenue was $59.3 billion in Q4, up 27%, and commercial remaining performance obligation increased 84% to $678 billion.
Satya Nadella, chairman and CEO of Microsoft, said Azure revenue surpassed $100 billion for the first time and Microsoft 365 Copilot reached more than 30 million paid seats. He said those milestones reflect customer confidence in Microsoft’s role powering AI transformation.
The quarter gives Microsoft a stronger proof point for its AI platform strategy. The company is not only selling cloud capacity. It is arguing that enterprise AI will depend on model choice, governed data, context, agent controls, developer platforms, and business applications working together.
Multi-Model AI Becomes the Cloud Battleground
On Microsoft’s earnings call, Nadella said the company now offers more than 11,000 models across Azure, including models from OpenAI, Anthropic, Mistral, xAI, and Microsoft’s own MAI family. He also said Microsoft has seen a fivefold increase since the start of the year in customers building with models from multiple providers.
That is a critical signal for enterprise software buyers. AI strategy is moving away from single-model bets and toward model portfolios selected by quality, latency, cost, compliance, and business continuity. Nadella said Microsoft is building a system where the harness, context, memory, and action space are separated from any one model family, making models substitutable.
For ERP leaders, that architecture matters because business applications will increasingly use different models for different tasks. A finance agent, customer-service agent, developer assistant, sales workflow, or supply-chain process may not need the same model. The platform question becomes where those models are governed, how they access enterprise data, and how agent actions are controlled.
Microsoft pointed to Levi Strauss & Co. as an example, saying the company is using OpenAI and Anthropic models on Foundry as it brings more than 1,000 domain-specific agents into a unified enterprise AI platform.
Dynamics, GitHub Show Where AI Turns into Workflow
Microsoft’s business applications story is also becoming more agentic. Dynamics 365 revenue increased 13% in Q4, or 12% in constant currency, while Nadella said Microsoft is exposing more than 650,000 MCP actions across sales, finance, supply chain, HR, and customer service.
Those actions allow agents to access business context and take action using the same data models, rules, permissions, security guardrails, and audit trails as application users.
That reinforces Microsoft’s broader MCP strategy. The company is turning Model Context Protocol into a connection layer between agents and business systems, including ERP and CRM workflows. For customers, the promise is that agents can act inside governed application structures rather than operate as disconnected automations.
GitHub adds another dimension to the story. Nadella said GitHub Copilot now has 50 million users and that GitHub has 225 million users overall, with more than 90% of the Fortune 500 using the platform for AI-powered development. He also said Copilot revenue accelerated more than 60% quarter over quarter after Microsoft introduced usage-based billing.
The implication is that Microsoft’s AI business model is moving beyond seats alone. Microsoft is layering consumption-based economics into Copilot, Cowork, Dynamics 365, GitHub Copilot, and future agentic offerings, tying revenue more closely to usage and outcomes.
What This Means for ERP Insiders
Enterprise AI is becoming multi-model by default. Microsoft’s 11,000-model catalog and fivefold increase in multi-provider customers show that buyers want flexibility across model families, cost profiles, latency needs, and compliance constraints. For ERP architects and AI platform teams, the next design priority is deciding how model choice will be governed across finance, supply chain, HR, sales, service, and development workflows.
Agentic ERP will depend on governed actions, not just smarter answers. Microsoft’s Dynamics 365 MCP actions show how agents are being connected to business context, permissions, audit trails, and application rules. For CIOs, CFOs, and process owners, the practical test is whether agents can execute work inside existing controls rather than create a new layer of untracked automation.
Cloud growth is now tied to AI operating models. Azure’s $100 billion annual revenue milestone and Microsoft Cloud’s Q4 growth show that enterprise AI demand is expanding the role of hyperscalers beyond infrastructure. For ERP vendors, systems integrators, and enterprise buyers, cloud strategy now needs to account for model governance, agent deployment, data grounding, usage economics, and cross-application execution.
