Air offers a platform for orchestrating numerous coding agents in parallel, while Junie CLI functions as a versatile coding agent compatible with large language models from Anthropic, OpenAI, Google, and more.
<div class="media-with-label__label">
Credit: Photography by VladyslaV Travel / Shutterstock </div>
</figure>
</div>
</div>
</div>
</div>
<div id="remove_no_follow">
<div class="grid grid--cols-10@md grid--cols-8@lg article-column">
<div class="col-12 col-10@md col-6@lg col-start-3@lg">
<div class="article-column__content">JetBrains has unveiled two innovative tools aimed at enhancing AI-powered software development: Air, a platform designed for concurrently managing and deploying multiple AI coding agents, and Junie CLI, a coding agent engineered to be compatible with various large language models (LLMs).
Both products were announced on March 9. Air, now available in public preview, can be accessed and downloaded from air.dev. Meanwhile, Junie CLI, currently in beta, is available for use at junie.jetbrains.com.
Air, presently offered free for MacOS users with upcoming versions for Linux and Windows, is an agentic development environment (ADE). It’s conceptualized around integrating essential tools for managing coding agents into a cohesive user experience, as stated by JetBrains. Functioning as a unified workspace where agents like Claude, Gemini CLI, Codex, and Junie CLI can operate in tandem, Air empowers developers to navigate their codebase efficiently and seamlessly switch between different coding agents. When assigning a task, developers can specify a particular line, commit, class, method, or other symbol, thereby providing the agent with precise context rather than a vague block of text. Upon task completion, Air presents the modifications within the broader context of the codebase, complemented by crucial tools such as a terminal, Git integration, and a built-in preview, according to JetBrains. Future plans for Air include expanding support for additional coding agents via the Agent Client Protocol (ACP) through the ACP Agent Registry, the company confirmed.
Similar to Air, Junie CLI is engineered to ensure that agent-generated code is robustly integrated within the existing codebase. This standalone coding agent is designed to be LLM-agnostic, supporting a wide range of high-performing models. It is built to tackle complex challenges, maintain context awareness by default, and guarantee reliability and security, JetBrains explained. With its anticipated March release, Junie CLI will be available for direct use from the terminal, within any IDE, in CI/CD pipelines, and on platforms like GitHub or GitLab. Junie CLI currently supports leading models from OpenAI, Anthropic, Google, and Grok, with plans to integrate new models as they become available.
