JetBrains just launched Air, and it might change how you work with AI coding agents forever.
After 26 years of building IDEs like IntelliJ IDEA and PyCharm, JetBrains is now tackling the next frontier: an environment built specifically for working with multiple AI agents. And it is free to try right now.
Let me break down what Air is, why it matters, and whether you should care.
What is JetBrains Air?
Air is an agentic development environment — not a chat window, not a code editor with AI sprinkled on top. It is designed from the ground up to orchestrate multiple AI agents working on your codebase simultaneously.
Think of it this way: Your current setup probably has different agents scattered across different tools. One agent in your terminal. Another in your IDE. A third in a browser tab. Each one has its own context, its own setup, and zero understanding of what the others are doing.
Air brings them all together. You can run Codex, Claude Agent, Gemini CLI, and Junie in one place, switch between them freely, and run tasks concurrently without the chaos.
The Problem Air Solves
If you have been using AI coding agents for a while, you know the pain points:
- Fragmented context: Each agent sees your codebase differently (or not at all)
- Tool switching fatigue: Jumping between windows, terminals, and browser tabs
- No structural understanding: Agents work on text blobs, not code structures
- Review chaos: Diff views that show changes without project context
Air addresses all of this by doing something smart: it gives agents precise context instead of pasted text. You can mention a specific line, commit, class, or method, and the agent understands exactly what you mean.
When you review the output, you see changes in the context of your entire codebase — with a terminal, Git client, and preview tools built right in.
Multi-Agent Orchestration
This is where Air gets interesting.
You can run multiple agents at the same time, each working on different tasks. Air isolates them in Docker containers or Git worktrees, so they do not stomp on each other's changes.
The UI shows one task at a time, and you get notified when another task needs attention. Switch over, handle it, switch back. The agent keeps working in the background.
For teams working on larger codebases, this concurrent approach is genuinely useful. You can have one agent refactoring authentication while another writes tests for a new feature — all in parallel, all isolated.
Supported Agents
Right now, Air supports:
- OpenAI Codex
- Claude Agent
- Gemini CLI
- Junie
It uses the Agent Client Protocol (ACP), which means more agents will likely be added soon through the ACP Agent Registry.
The best part? You can switch agents between projects naturally. No migration, no configuration headache. Just pick the agent you want for the task.
Pricing and Access
Here is the good news: Air is free for JetBrains AI subscribers.
If you have:
- JetBrains AI Pro (included in All Products Pack and dotUltimate)
- JetBrains AI Ultimate
You can just sign in and go. All agents are included.
Prefer to use your own API keys? You can bring your own keys from OpenAI, Google, or Anthropic. Air will use your keys first, then fall back to the JetBrains subscription for anything not covered.
A dedicated enterprise offering is coming soon, with cloud execution for remote agent runs.
Why This Matters
The AI coding space is moving fast. Every week brings a new model, a new agent, a new tool promising to revolutionize development.
What JetBrains understands — after 26 years of building developer tools — is that integration matters more than novelty. A fancy new agent is useless if it does not understand your codebase structure, your Git history, your project dependencies.
Air solves the real problem: giving agents the context they need to be genuinely useful, without making developers jump through hoops.
Should You Try It?
Yes, if:
- You use multiple AI coding agents regularly
- You work on complex codebases that need structural understanding
- You want to run agents in parallel without manual sandboxing
- You already have a JetBrains subscription
Maybe wait, if:
- You only use one agent and are happy with your current workflow
- You need enterprise features (coming soon)
- You are on Windows or Linux (currently macOS only, though other platforms are coming)
Getting Started
- Download Air from the JetBrains website (free for macOS, Windows and Linux coming soon)
- Sign in with your JetBrains Account
- Connect your preferred agents or bring your own API keys
- Define your first task and watch it work
If you want to go deeper into agentic development and AI-assisted coding, I recommend checking out Clean Code by Robert Martin — it is the foundation for writing code that AI agents can actually help you improve. For a productivity framework that pairs well with AI tools, Deep Work by Cal Newport is essential reading.
The Bottom Line
JetBrains Air is not just another AI tool. It is an attempt to solve the orchestration problem that every developer working with agents faces.
The agents are getting better. The models are getting smarter. But without proper context and coordination, they are still just fancy autocomplete.
Air gives them structure. And that might be the piece missing from your AI workflow.
Disclosure: Some links in this post are affiliate links. As an Amazon Associate, I earn from qualifying purchases. This helps support the content at no extra cost to you.
Top comments (0)