OpenAI Launches New macOS App for Codex, Integrating AI-Driven Coding Agents

According to the Economic Desk of Webangah News Agency, OpenAI has taken a major leap in AI-driven software development with the launch of its new macOS application for Codex. The app integrates popular agent-based workflows developed over the past year, enabling parallel operation with multiple AI agents and advanced workflow capabilities. This release comes less than two months after OpenAI introduced GPT-5.2-Codex, its most powerful coding model to date.

Sam Altman, CEO of OpenAI, emphasized during a press conference that while GPT-5.2 is currently the most capable model for complex tasks, its usability has been challenging. The new macOS app aims to address this by combining the model’s capabilities with a more flexible user interface.

Benchmark tests present a nuanced picture of GPT-5.2’s performance. While it maintains top position in TerminalBench for AI programming tasks, competitors like Gemini 3 and Claude Opus show comparable results. The SWE-bench, which evaluates AI’s ability to fix real-world software bugs, similarly shows no clear advantage for GPT-5.2.

The Codex application introduces several new features that OpenAI claims will help it match or surpass various cloud applications in certain areas. These include customizable automation that can run in the background on a schedule, with results queued for user review. Users can also select different AI agent personalities to match their working style.

OpenAI highlights the accelerated development pace enabled by AI as the strongest selling point of its new offering.