OpenAI has made its Codex AI coding agent available directly inside JetBrains IDEs, enabling developers to plan, write, test, review and deploy code without leaving their editor. OpenAI and JetBrains announced this move on January 22.

With Codex now natively integrated into JetBrains AI chat, developers now can collaborate with Codex directly within a JetBrains IDE, starting with version 2025.3. Codex can be used with a JetBrains AI subscription, a ChatGPT account, or an OpenAI API key, all within the same AI сhat interface. Specific IDEs supporting this capability include JetBrains’s IntelliJ, PyCharm, WebStorm, and Rider.

Codex in JetBrains IDEs is powered by the GPT-5.2-Codex coding model from OpenAI. This model can undertake complex tasks such as large refactors, code migrations, and feature builds. GPT-5.2-Codex is positioned as OpenAI’s most advanced coding model, with significantly stronger cybersecurity capabilities, helping teams find vulnerabilities, reason about exploits, and harden security systems at scale, according to OpenAI.

Using Codex, developers can delegate real coding tasks from within an IDE and let the agent reason, act, and iterate alongside them. Codex supports various interaction modes, with users able to decide how much autonomy to give it—from simple question-response permissions to the ability to access a network and run commands autonomously. Users can switch between supported OpenAI models and their reasoning budget directly in the AI chat, making it easy to balance reasoning depth, speed, and cost depending on the task at hand.

The Codex agent is available for free for a limited time when accessed via JetBrains AI, including the free trial or free tier version. The promotion starts on January 22 and will remain available until allocated promotional credits have been used up. This free offer does not apply when using a ChatGPT account or an OpenAI API key, JetBrains said.