AI TL;DR
GPT-5.2-Codex is now built into IntelliJ, PyCharm, and more. Here's how to set it up and get the most out of AI coding assistance. This article explores key trends in AI, offering actionable insights and prompts to enhance your workflow. Read on to master these new tools.
OpenAI Codex in JetBrains: Complete Setup Guide
OpenAI and JetBrains announced native Codex integration on January 22, 2026.
Now you can use GPT-5.2-Codex—OpenAI's most advanced coding AI—directly in IntelliJ, PyCharm, WebStorm, and Rider. Here's how to set it up.
What Is GPT-5.2-Codex?
The Coding-Optimized Model
GPT-5.2-Codex is OpenAI's specialized coding variant:
| Feature | Description |
|---|---|
| Optimized for code | Better understanding of programming contexts |
| Agentic capabilities | Can plan, execute, and iterate on tasks |
| Large refactors | Handles complex multi-file changes |
| Code migrations | Assists with language/framework upgrades |
Difference from Regular GPT-5.2
| GPT-5.2 | GPT-5.2-Codex |
|---|---|
| General purpose | Code-optimized |
| ChatGPT app | IDE integration |
| Conversation-based | Task-based |
| Broad knowledge | Deep programming focus |
Setup Requirements
What You Need
-
JetBrains IDE version 2025.3 or later
- IntelliJ IDEA
- PyCharm
- WebStorm
- Rider
-
AI Assistant Plugin (latest version)
-
One of these access methods:
- JetBrains AI subscription
- OpenAI API key
- ChatGPT account (Plus or higher)
Step-by-Step Setup
Method 1: JetBrains AI Subscription
- Open Settings (Ctrl+Alt+S or Cmd+,)
- Navigate to Plugins → Marketplace
- Search "AI Assistant" and install/update
- Go to Settings → Tools → AI Assistant
- Sign in with JetBrains account
- Select Codex as your coding model
- Done! Use via AI chat or right-click menu
Method 2: OpenAI API Key
- Get API key from platform.openai.com
- Open Settings → Tools → AI Assistant
- Choose "OpenAI" as provider
- Paste your API key
- Select "GPT-5.2-Codex" model
- Save and start coding
Method 3: ChatGPT Account
- Open Settings → Tools → AI Assistant
- Choose "Sign in with ChatGPT"
- Authenticate via browser
- Return to IDE (automatically detected)
- Verify Codex access in settings
Using Codex in Your IDE
AI Chat Panel
Access via: View → Tool Windows → AI Assistant
Use cases:
- "Explain this function"
- "Find bugs in this code"
- "Refactor to use async/await"
- "Write unit tests for this class"
Inline Actions
Right-click on any code selection:
- Generate → Create code from description
- Explain → Get explanation of selected code
- Refactor → AI-assisted refactoring
- Fix → Suggest fixes for errors
- Test → Generate test cases
Commit Assistance
When committing:
- Auto-generate commit messages
- Summarize changes
- Check for issues before push
Best Practices
Prompting Tips
| Good | Less Effective |
|---|---|
| "Refactor this to use dependency injection" | "Make this better" |
| "Add error handling for network failures" | "Add error handling" |
| "Convert this Python 2 code to Python 3" | "Update this code" |
Context Matters
Codex works better when:
- ✅ File is open in editor
- ✅ Related files are in project
- ✅ You select specific code sections
- ✅ You provide clear context in prompts
For Large Refactors
- Start small — Test on one file first
- Review changes — Always verify before applying
- Use version control — Commit before major refactors
- Iterate — Break large tasks into smaller ones
Workflow Examples
Example 1: API Migration
Prompt: "Migrate all REST calls in this file from Retrofit to Ktor"
Codex will:
1. Identify Retrofit calls
2. Generate Ktor equivalents
3. Update imports
4. Handle response parsing
Example 2: Test Generation
Prompt: "Write unit tests for the UserService class
covering all public methods"
Codex will:
1. Analyze UserService
2. Create test class structure
3. Generate test cases
4. Include edge cases
Example 3: Bug Investigation
Prompt: "This function throws NullPointerException
sometimes. Find and fix the issue."
Codex will:
1. Analyze null paths
2. Identify risky operations
3. Suggest defensive fixes
4. Explain the root cause
Pricing
JetBrains AI Pro
- Included with JetBrains All Products subscription
- Or standalone AI subscription
- Unlimited Codex usage within quota
OpenAI API
- Pay-per-token pricing
- GPT-5.2-Codex rates apply
- Usage tracked in your account
ChatGPT Account
- Requires Plus ($20/mo) or higher
- Shared quota with ChatGPT usage
- Convenient for existing subscribers
Tips for Code Quality
Review AI Output
Always check:
- ✅ Logic correctness
- ✅ Security implications
- ✅ Performance impact
- ✅ Code style consistency
- ✅ Edge case handling
Iterate Collaboratively
Think of Codex as a pair programmer:
- Initial suggestion → Your review → Refinement request
- Don't accept everything blindly
- Guide it toward your standards
Troubleshooting
Common Issues
| Problem | Solution |
|---|---|
| Codex not appearing | Update AI Assistant plugin |
| Authentication fails | Check account permissions |
| Slow responses | Check internet connection |
| Wrong language output | Specify language in prompt |
| Context issues | Open relevant files in editor |
Our Take
This integration makes AI-assisted coding seamless.
Before: Copy code → Paste in ChatGPT → Copy response → Paste back Now: Select code → Right-click → Apply fix
For developers who live in JetBrains IDEs, this is a significant workflow improvement. The GPT-5.2-Codex model is genuinely capable of complex refactoring and understanding multi-file contexts.
That said, always review AI-generated code. It's a powerful assistant, not a replacement for developer judgment.
Are you using AI coding assistants? How does Codex compare to Copilot or Cursor?