On November 3, 2026, GitHub and Microsoft detailed Copilot in JetBrains Air’s agent picker, including local IDE tasks and optional Air Teams cloud execution for eligible organizations. It connects an existing Copilot subscription to Air’s agent workspace. Cloud execution still requires Air Teams access and administrator approval. Developers can work locally in their IDE; eligible business teams can also delegate work to Air’s cloud environment.
What the illustration shows — and what it does not
The two ideas in the artwork are the local agent, which works in your IDE, and the cloud task, which runs remotely for eligible Air Teams organizations. You still choose the agent and review its changes.
What Happened
The November 3 GitHub release announcement explains how to select Copilot in JetBrains Air through the Agent Client Protocol (ACP), which connects coding agents to Air’s task and review workflow. Air can manage multiple agent sessions, present project context and IDE tools, and let a developer inspect results next to their code. GitHub also says the integration can run cloud tasks in isolated environments for eligible Air Teams organizations.
In July 2026, JetBrains documented running Copilot CLI through ACP in Air. The November 3 update described Copilot as an available default agent in Air and extended the workflow to Cloud Run for eligible Air Teams organizations. These are different stages of the integration. Air connects Copilot through its own agent picker, so that workflow does not require installing the separate Copilot plugin. The standalone plugin remains the broader option for inline assistance, code review and enterprise controls. A developer choosing between the two should compare how much task delegation, parallel review and team governance they actually need.
Who Is Affected & When
The November 3 release describes local use and eligible Air Teams cloud tasks as available; specific IDE and organization settings still determine access. For GitHub Copilot local coding tasks, GitHub instructs users to install Air in a JetBrains IDE, select Copilot in the agent picker, and connect their GitHub Copilot account. The official release does not establish a separate universal rollout date for every JetBrains IDE version, so compatibility should be checked against the installed Air plugin.
GitHub Copilot Air Teams cloud tasks have additional conditions: a JetBrains account in a business organization with Air Teams enabled, an existing GitHub Copilot subscription, and Copilot enabled by the Air administrator. Individual users should not assume that a personal Copilot plan alone unlocks the cloud workflow. GitHub’s announcement does not promise cloud-task availability in every organization or region.
What Changes in Practice
| Requirement | Local work in Air | Cloud work with Air Teams |
|---|---|---|
| Who can use it | Developers with a Copilot subscription and the Air plugin. | Members of an Air Teams-enabled business organization, with Copilot enabled by its administrator. |
| Where tasks run | In the local JetBrains development environment. | On a dedicated cloud VM, with the agent in a Docker container and network traffic routed through a proxy. |
| How to review work | Inspect changes in Air alongside the local project. | Start from the IDE or browser and review the completed cloud task in Air. |
| Billing visibility | Check Copilot subscription and usage with GitHub. | Copilot usage is not included in Air’s governance analytics; its spending limits cannot be set in Air. |
Local tasks
- Install the Air plugin in your JetBrains IDE.
- Choose GitHub Copilot from Air’s agent picker and sign in with your Copilot subscription.
- Start a local coding task and review the resulting changes in Air.
Cloud tasks through Air Teams
- Confirm that your organization has Air Teams enabled and that an administrator has allowed Copilot.
- In the browser, connect GitHub under Settings → VCS Providers, select the repository and remote branch, then choose Copilot and send the task.
- Alternatively, in the IDE’s Air plugin, select the Cloud environment before assigning a task.
Before running cloud tasks against production code, check repository permissions, the chosen branch and which credentials the cloud environment can access. Review generated changes under your normal pull-request and branch-protection rules.
The JetBrains explanation of cloud agent sandboxing describes a fresh virtual machine with a container for each cloud task, plus a network proxy and restricted access to resources. This explains how Air isolates remote coding tasks; it does not establish whether the agent’s changes are correct or secure.
Why It Matters
For teams already using Copilot, Air may help coordinate local coding and cloud tasks without introducing a separate agent subscription. GitHub says Copilot tasks in Air use the Copilot plan rather than JetBrains AI credits. That is an important billing distinction, but it does not establish that all tasks are unlimited or free: users must consult their own subscription and organizational policies.
GitHub warns that subscription or credit availability may be checked when a cloud environment starts; an unavailable entitlement can cause the task to fail. Air administrators can disable Copilot through governance policy, but GitHub says they cannot configure Copilot spending limits inside Air. Teams that need hard budget controls should verify those controls with their billing administrator before scaling usage.
Enterprise cost reporting also has two gaps. GitHub states that users cannot choose, for each task, whether a personal or organizational Copilot plan is charged. Air’s governance analytics do not include Copilot usage. Those constraints can complicate cost allocation and internal reporting for businesses operating multiple subscriptions.
Our Take
For developers already paying for Copilot and using JetBrains tools, the JetBrains Air Copilot integration is worth evaluating when work involves several independent coding tasks, repeated reviews, or a team-managed cloud environment. For a team handling multiple independent coding tasks, the benefit is tracking agent runs and reviewing changes within one Air workspace. For a solo developer who mainly wants inline completions, the existing Copilot plugin may still be sufficient.
Start with a noncritical task. Verify which Copilot plan is charged, that the administrator has allowed the agent and that the review process works as expected. DigiOffly has not run a hands-on test of Air, provisioned a cloud task, measured execution speed or verified actual billing on a customer account.
GitHub Copilot and JetBrains Air FAQ
Is GitHub Copilot in JetBrains Air available to individual developers?
Yes. A developer with the supported Air plugin, IDE setup and Copilot plan can use local tasks. Cloud execution separately requires Air Teams eligibility and administrator approval.
Do Copilot tasks use JetBrains AI credits?
According to GitHub and JetBrains, this integration uses the existing GitHub Copilot plan rather than JetBrains AI credits. Applicable usage limits still depend on that Copilot plan.
Can a developer run Air cloud tasks without Air Teams?
The vendor documentation currently describes cloud execution for Air Teams-enabled business organizations. A personal Copilot subscription alone does not grant that cloud workflow.
Does JetBrains Air show Copilot spending or allow spending caps?
Not for Copilot usage. Air’s governance analytics exclude Copilot activity, and Air administrators cannot set Copilot spending limits within Air. Check the applicable controls in GitHub.
Official Sources & Dates
GitHub Changelog — GitHub Copilot is now available in JetBrains Air (announced November 3, 2026; availability described as current). JetBrains Air Blog — Use Your GitHub Copilot Plan in JetBrains Air (November 2026; setup and account guidance). Microsoft Developer Blog — Copilot as a Default Agent in JetBrains Air (November 3, 2026; separate plugin requirements and feature comparison). JetBrains Air — ACP agent support (July 2026) (background to the November announcement). JetBrains Air Blog — How Cloud-Based Agent Sandboxing Works in Air (October 2026; technical context and organizational availability). Availability and billing can change; check current vendor documentation before purchase or rollout.
For background on Copilot’s broader capabilities and limitations, see the DigiOffly GitHub Copilot review.