OpenAI launches usage analysis and spend control capabilities for ChatGPT Enterprise

OpenAI launches credit usage analysis and granular spend control capabilities for ChatGPT Enterprise. The new Global Admin Console integrates the credit usage of ChatGPT and Codex into a unified view, supports splitting consumption by users, products and models, and provides in-depth analysis capabilities through the Cost API. In terms of spending control, administrators can set default limits for workspaces, configure quotas for specific groups, and create individual coverages for those who need more capacity. Employees can also view their credit usage compared to their budget and apply for additional credits.

Zipcode co-founder Ryan Oksenhorn commented: "Zipcode's engineering team has been fully committed to Codex since January... We asked OpenAI to build usage analytics to help identify and train employees who have not yet adopted Codex, as well as granular usage controls to keep spend predictable. These new tools are helping us scale employee productivity while maintaining safety guardrails."

This is another important signal that enterprise-level AI is moving from "unlimited experimentation" to "refined operation" after Microsoft's Copilot Cowork shifted to pay-as-you-go billing. Unlike DeepSeek and other models whose core competitiveness is low price, OpenAI's strategy is to use management tools to allow customers to be budget-conscious even on high-priced models - this is actually a battle between two different enterprise AI business models: cost advantage vs. management sophistication.

It is worth paying attention in the future: whether this feature can accelerate the enterprise penetration of ChatGPT Enterprise, and whether domestic AI companies (Doubao, DeepSeek, Zhipu) will follow up with similar cost management tools.