OpenAI Codex officially lands on mobile! ChatGPT App can directly write code and review code

📄Full Article· Automatically extracted by trafilaturaGemini 翻譯1952 words

OpenAI announced that the Codex AI coding agent has officially landed on the ChatGPT mobile app. Developers can now initiate new tasks, review code outputs, and steer execution flows on iOS and Android, bringing professional development work to mobile and enabling a true mobile development experience. (Previous coverage: OpenAI launches new engineer agent Codex! AI can write features, fix bugs, run tests... limited to 3 types of users for early access) (Background: Codex mobile operation goes live: remotely supervise AI, start new tasks, and view outputs directly in ChatGPT) OpenAI announced via its official X account on May 21 (Taipei Time) that its AI coding agent tool, Codex, has officially landed on the ChatGPT mobile app, supporting both iOS and Android platforms. This update means developers no longer need to be tethered to a desktop to write and debug code; by simply opening their phones, they can start Codex sessions, review outputs, steer execution, and approve next steps at any time. According to OpenAI's official documentation, the mobile version of Codex is now available in all regions where ChatGPT is supported, with mobile connectivity for Windows users coming soon. This feature upgrades ChatGPT from a conversational chatbot to a true mobile engineering assistant, allowing developers to maintain control over their projects during commutes, between meetings, or even while traveling. Notably, this is not the first time Codex has intersected with mobile. In fact, OpenAI began laying the infrastructure for "development anywhere" earlier this year—from a Chrome extension that allows Codex to operate computers while logged in, to backend Mac control and embedded browsers, and now the ChatGPT App integration, this mobile development front is rapidly taking shape. You've been asking for this one… Now in preview: Codex in the ChatGPT mobile app. Start new work, review outputs, steer execution, and approve next steps, all from the ChatGPT mobile app. Codex will keep running on your laptop, Mac mini, or devbox. pic.twitter.com/9i2Jckjt9z — OpenAI (@OpenAI) May 14, 2026 The Era of Mobile Development Agents: The Game Between Codex and Competitors With OpenAI Codex landing on mobile, its most direct competitor is the app version of Anthropic Claude. While Claude also supports mobile code completion and analysis, its positioning is more of a general-purpose AI assistant. In contrast, Codex was built from the ground up as an engineering agent specifically for software development—it can run tests, fix bugs, and even automatically submit Pull Requests. On the other hand, Grok, the AI assistant launched by xAI, possesses advantages in real-time web information and X platform integration, but still has a clear gap in pure programming capabilities compared to Codex. Cursor, which starts from the IDE, takes a different path—it is a desktop-first development environment integration, making it difficult to replicate the mobile-native experience of the ChatGPT App in the short term. From the perspective of a developer's workflow, the greatest significance of Codex on mobile is breaking the traditional boundary that "development = sitting in front of a computer." In the past, mobile-related tools were mostly for reading files, lightweight editing, or monitoring CI/CD status; agent tools capable of truly "executing" and "steering" development work were never popularized. Codex's approach moves the entire execution loop of an AI agent—starting tasks, executing, reviewing outputs, steering the next steps, and approving completion—into the phone, allowing developers to maintain control over code quality and progress even when away from their desks. Reactions and Opportunities in the Local Developer Community The local software developer community has reacted enthusiastically to this update. In several major technical groups, developers have already shared their hands-on testing experiences: using Codex on mobile to review GitHub Issues, having the AI generate fix code, and then approving it after reviewing it through the ChatGPT App. "I used to only be able to reply to Slack on the MRT, but now I can let Codex run the test suite first and see the results as soon as I get to the office, effectively turning commute time into development time," a Taipei-based backend engineer noted in a discussion. Developers in the cybersecurity field also pointed out that the output review function of Codex on mobile is well-designed—AI requires manual approval from the developer before performing any operation, striking a good balance between convenience and security. For the thriving local startup ecosystem, the mobile version of Codex may bring more direct benefits: startup teams have limited resources, and developers often have to wear multiple hats. The arrival of the mobile version of Codex allows founders to outsource routine development work to an AI agent without increasing personnel costs, enabling them to focus on product strategy and business development.

Data Status✓ Full text extractedRead Original (動區 BlockTempo)

🔍Historical Similar Events· Keyword + Asset Matching6 items

2026-05-16

OpenAI Consolidates Product Lines! Greg Brockman Simultaneously Takes Over ChatGPT, Codex, and the Developer API

2026-05-15

OpenAI confirms two employee computers were affected by the TanStack supply chain attack; ChatGPT and Codex macOS users must update before 6/12

2026-05-15

OpenAI launches "ChatGPT Personal Finance" feature: Connects to real bank accounts to help you calculate expenses and plan mortgages

Similarity 170%關鍵字 chatgpt/openai同分類 zh

2026-05-15

Codex mobile operations go live: directly supervise AI remotely on ChatGPT, launch new tasks, and review output results

Similarity 170%關鍵字 chatgpt/codex同分類 zh