JetBrains Air Governance is now generally available for organizations

One console for several AI vendors

Air Governance is the organization layer of JetBrains Air, which JetBrains describes as a system of products for agentic development. According to the company, it works with Claude Agent, Codex, Gemini CLI, Junie and other ACP-compatible agents, across IDEs, the CLI, web browsers and mobile.

JetBrains says the problem it addresses is common: most engineering teams now use several AI tools from different providers at once. Each tool has its own contract, dashboard and admin settings, which makes it hard to see who uses which agents and models, how much AI costs the company and whether it helps.

The company says it faced the same issue internally. Its AI spend grew roughly tenfold in the first half of 2026, and its first cross-vendor usage report took four days to compile by hand. The tools built in response became Air Governance, which JetBrains says now helps manage the agentic usage of roughly 2,300 people at the company every month.

Because the console works across vendors, developers can pick the agents and models that suit their tasks. When a new agent or model appears, it can be added to the same policies, with no new contract or separate admin setup, according to JetBrains.

Controls and cost tracking

In the console, admins can:

• Set AI credit limits for individual users, teams and service accounts, so budgets are enforced rather than just reported.
• Manage AI providers and set up centralized BYOK, using JetBrains AI or connecting their own OpenAI, Anthropic, Amazon Bedrock or Google Vertex AI accounts.
• Define which AI agents are approved across the organization.
• Configure MCP servers at the company level, so everyone uses the same approved external tools.
• Create AI policies that combine approved providers, models, agents and credit limits for different user groups.

On the reporting side, Air Governance shows AI costs by user, by agent, and by model and provider. This lets organizations compare spend across Junie, Claude Agent, Codex, Gemini CLI and other agents, and see which models drive the bill.

It also offers AI outcome metrics meant to show how AI changes the way developers work. These include the split between human-authored and AI-generated code, and how much code is AI-generated or AI-modified. Details on how each metric is calculated are in the AI analytics documentation.

Pricing, roadmap and what it means for teams

Air Governance is available to all JetBrains organizational customers, with free and paid plans. It is included in the Business plan, which costs USD 20 per month for each developer with AI access. Developers do not need individual AI licenses. They draw on a shared pool of AI credits, which the organization buys separately and which cover usage across JetBrains products and connected agents.

JetBrains lists several capabilities it is working on:

• MCP server propagation to third-party CLIs.
• GitHub Copilot support through Air Gateway, its connector for third-party CLIs.
• Connections to cloud agents through Air Gateway.
• Advanced AI analytics.
• Credit limits and custom pricing for BYOK models.
• The ability to manage skills across teams.

For companies running a mix of AI coding tools, the release offers a single place to apply budgets and approvals instead of handling each vendor separately. Existing customers can open Air Governance in the console, which requires admin access. JetBrains is also holding a webinar on November 18, 4:00 to 5:00 PM CET, covering access and usage policies, spending limits and analytics.