Microsoft Replaces Flat AI Billing with Usage-Based Model

Microsoft is transitioning GitHub Copilot from a flat-rate, request-based billing system to a usage-based model effective June 1, 2026. This decision aims to alleviate financial strain resulting from increasing inference costs associated with providing AI services. The new system will correlate prices directly with metered tokens, which reflect the complexity and resource demand of user requests.

The implications of this change are significant for the landscape of AI service pricing. By adopting a usage-based approach, Microsoft seeks to ensure sustainability in operating costs while enhancing transparency for users. However, this pivot may lead to an increased cost for developers relying on GitHub Copilot, potentially prompting the need for alternative AI solutions and raising questions about dependency on external AI capabilities.