Microsoft Halts Tokenmaxxing! Budget Capped, Employees Bear Overages
Microsoft has adjusted its internal AI usage strategy, halting Tokenmaxxing behavior and setting budget caps. The default model is now GPT-5.6, and any costs exceeding the limit will be borne by employees.
Microsoft recently adjusted its internal policy on the use of AI tools, explicitly halting high-frequency calling behavior known as Tokenmaxxing. The company has set strict budget caps, and costs exceeding this limit will be borne by employees themselves, aiming to control the operational costs of internal AI applications.
This adjustment involves the selection of the internal default model, which will uniformly adopt the GPT-5.6 version. This move reflects that large tech companies, in the process of advancing AI implementation, are beginning to shift from simply pursuing functionality to focusing on the balance between resource consumption and cost-effectiveness.
As the capabilities of large models improve, the computational cost per interaction is also increasing. Microsoft's move may signify that enterprise-level AI applications are entering a stage of fine-grained operations, avoiding computing power waste caused by disorderly calling, and ensuring the technology ROI meets expectations.
For the industry, this policy change serves as a signal. Other tech giants may refer to similar cost control strategies, establishing more comprehensive usage monitoring and budget management mechanisms while encouraging employees to use AI tools.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.