How JetBrains AIR Redefines IDEs with Built‑In AI: A Hands‑On Guide

What Is AIR?

JetBrains AIR is a brand‑new, standalone IDE built from the ground up with AI as the default workflow rather than an optional add‑on. Its core principles are:

AI is the default work flow, not a supplemental feature.

Emphasis on determinism and explainability to avoid “black‑box magic”.

Developers retain full control over the generated code.

Quick Start: Three Steps to Experience AIR’s AI‑Native Flow

1. Install and First Run

Currently AIR supports only macOS; Windows and Linux versions are planned for 2026. After downloading, launch the application, sign in with a valid subscription, and you will be taken to the welcome screen.

2. Open Project

There are two ways to open a project:

Open a local project: click Open and select the folder.

Clone a Git project: click Clone from Git , enter the repository URL and the destination path.

When a project is opened for the first time, AIR asks whether to trust the code. Options are:

Preview : limited functionality, code cannot be executed.

Trust : allows code execution (recommended only for trusted projects).

3. Define Task

Interact with the AI agent via a chat‑like interface to gradually clarify the task. Provide a clear description (e.g., “Add pagination to the order module, 10 items per page”) and optional context such as a file path (e.g., “refer to src/orders/service.py”).

Two interaction modes are supported:

Plan mode : the AI first generates an execution plan, then refines the task step by step.

Task configuration also includes permission settings:

Ask for permission on first use.

Automatic editing: accept file edits automatically.

Planning mode: only analyze code without editing or executing.

Full access: skip all permission prompts.

Context can be added in several ways:

Files and folders.

Git branches / commits / local changes.

MCP server.

Terminal tabs.

Upload local files.

Select code snippets directly.

Running and Task Management

Task states include:

Running : AI is executing; you can provide additional input at any time.

Waiting for user action : AI pauses and requires manual intervention.

Finished : task completed and ready for review.

Canceled / Archived : manually stopped or archived after exit.

Multiple tasks can run in parallel (e.g., one task writes tests while another fixes a bug) without interfering with each other.

Review and Submit Changes

In the Review tab, the diff panel shows AI‑generated modifications. Two view modes are available:

Unified diff: old and new code on the same page.

Split diff: side‑by‑side comparison.

You can add comments next to line numbers, just like a regular code review.

Advanced Feature: MCP Server Integration

Through the Model Context Protocol (MCP) you can connect external tools such as databases or APIs. Configuration is done by pasting a JSON block into the settings (Settings → AI → MCP Servers).

{
  "mcpServers": {
    "alphavantage": {
      "command": "uvx",
      "args": ["av-mcp", "YOUR_API_KEY"]
    }
  }
}

The example above shows how to connect to the Alpha Vantage API.

Conclusion

JetBrains AIR lowers the barrier for AI‑assisted programming by providing a clear, visual workflow: define the task with context, configure execution environment and permissions, then review and commit the changes. Parallel task execution, MCP extensions, and local snapshots further enhance efficiency and control for complex projects.