JetBrains Details Its First Steps to Bring Rapidly Growing AI Spend Under Control

JetBrains has outlined its initial steps to manage its rapidly increasing AI spending. In the past six months, development-related costs nearly quadrupled. Rather than limiting engineers to a restricted set of approved tools, JetBrains created a shared access and accounting layer to maintain tool flexibility while improving visibility and control over consumption.

As AI usage surged, most developers were utilizing three to five AI tools per month. By January 2026, token consumption began to climb sharply following the introduction of advanced models like Claude Opus 4.5 and 4.6. Initially, JetBrains manually compiled usage and expense data into spreadsheets over four days, which provided a snapshot of spending across providers and tools.

The company then automated the process using provider APIs and internal dashboards, offering teams a real-time view of current and forecasted expenditure. However, visibility alone did not enable intervention. Dashboards displayed where spending was happening, but requests still went directly from individual tools to their respective providers.

JetBrains developed Central CLI, a common command-line interface that serves as a unified entry point for both internal and external AI tools. By routing CLI requests through JetBrains' existing AI platform, the company introduced a centralized control point between developers and model providers, transforming the platform from a passive observer of AI spend to an active participant in traffic management.

The platform integrates JetBrains' existing AI-credit system, allowing managers to set spending limits for individual developers, teams, or broader groups. Unlike some companies that limit tool selection to a small number, JetBrains maintains a wide variety of supported options, understanding that the rapid pace of AI development makes it challenging to predict which tool will be most suitable for a task even a few months ahead.

Central CLI has been adopted by over 1,000 developers within weeks, although it does not yet cover all sources of AI expenditure. Some terminal-based agents and personal subscriptions remain outside the managed system, and the company is still developing policies for equitable AI budget distribution among users and teams. JetBrains' approach differs from others that have imposed restrictions on tool choice, such as Accenture's call for employees to reduce unnecessary AI usage or Uber's implementation of monthly AI budget limits.