In this video, I take a deep dive into the newly launched OpenAI Codex App for macOS. I explore its transition from a terminal/VS Code extension to a graphical interface, testing its capabilities in automation, skills, and project management while highlighting some baffling design choices and bugs.

--
Key Takeaways:

🖥️ OpenAI’s new Codex App offers a native-feeling macOS interface with background task capabilities.
🆓 Free plan users get one month of access, while subscribers enjoy 2x rate limits.
🤖 The app introduces "Skills" and "Automations" allowing for scheduled tasks and self-running agents.
🐛 The UI feels "vibecoded" with significant padding issues and visual glitches in the Skills tab.
🔒 Currently, model selection appears bugged, locking users to GPT-5.2 Codex Medium.
📉 Context awareness is inconsistent, sometimes requiring config deletions to read files properly.
🆚 The agentic UI falls short compared to competitors like Verdent, particularly in memory management and workflow.

AICodeKing

OpenAI launched a desktop app for Codex (GPT-5.2), available free for a month and currently macOS-only. The app provides a graphical interface with features like skills, automations, and work trees, drawing comparisons to Conductor. However, the review highlights numerous UI/UX issues including inconsistent design, buggy interfaces, poor context handling, unintuitive controls (like the plan mode toggle), and problematic VS Code integration that spawns multiple instances. The reviewer suggests competitors like Verdant offer superior agentic interfaces despite OpenAI's resources.

Codex Desktop App + Free GPT-5.2 Codex (Tested): Is OpenAI now copying Conductor,Commander?