OpenAI has released a new desktop application for Codex, aiming to streamline the management of multiple AI agents working concurrently on software projects. This macOS-based interface represents a significant evolution in AI-assisted development, moving beyond mere code generation to encompass comprehensive workflows that include design, building, testing, and maintenance.
The Codex app serves as a command center for developers, providing a centralized environment to run, monitor, and review tasks from various agents simultaneously. Each agent operates independently in its own thread, organized by project, enabling developers to easily switch between tasks without losing context.
OpenAI emphasizes the app’s ability to support isolated worktrees, allowing multiple agents to collaborate on the same repository without conflicts. Developers can review changes, comment on diffs, and open work directly in their editor, while also permitting agents to continue executing tasks independently without altering local Git states. Notably, session history and configuration are synchronized across the Codex command-line interface and IDE extension, facilitating a seamless transition for existing users to the new desktop app.
Since the initial launch of Codex in April 2025, OpenAI reports a shift in developer workflows, with a growing tendency to delegate substantial, long-running tasks to agents. The focus, the company states, is now centered on how humans supervise, direct, and trust these AI agents at scale.
In addition to code generation, the Codex app introduces a feature termed “skills,” which bundle instructions, tools, and scripts that allow agents to perform structured tasks such as research, documentation, deployment, and content creation. This functionality enables Codex to integrate with external services and workflows, including design tools and cloud platforms. OpenAI highlights that teams can create, manage, and share skills across projects, fostering a more consistent delegation of tasks that would otherwise be cumbersome to define repeatedly.
The app also features Automations, enabling agents to execute scheduled tasks in the background, including issue triage, testing, reporting, and documentation updates. Completed tasks are aggregated in a review queue, ensuring that human oversight is maintained without the need for constant supervision.
The Codex app is now available on macOS for users on ChatGPT Plus, Pro, Business, Enterprise, and Edu plans, with no additional fees for existing subscribers. For a limited time, it is also accessible to ChatGPT Free and Go users, who will benefit from increased rate limits across paid plans.
OpenAI reports rapid growth in Codex usage since the introduction of GPT-5.2-Codex in December, with over one million developers employing the system in the past month alone. The company plans to expand the app’s availability to Windows while continuing to refine multi-agent workflows, aiming to bridge the gap between the capabilities of advanced models and their practical application in the real world.
Meanwhile, the ETIH Innovation Awards 2026 have opened for submissions, recognizing education technology organizations that demonstrate measurable impact in K–12, higher education, and lifelong learning. This initiative invites entries from the UK, the Americas, and internationally, with assessments based on evidence of outcomes and real-world applications.
See also
Andrew Ng Advocates for Coding Skills Amid AI Evolution in Tech
AI’s Growing Influence in Higher Education: Balancing Innovation and Critical Thinking
AI in English Language Education: 6 Principles for Ethical Use and Human-Centered Solutions
Ghana’s Ministry of Education Launches AI Curriculum, Training 68,000 Teachers by 2025
57% of Special Educators Use AI for IEPs, Raising Legal and Ethical Concerns
