OpenAI's Codex Gains Desktop Control, Escalating Rivalry with Anthropic
The competition for the most effective AI coding assistant is intensifying. While Anthropic's Claude Code has recently gained favor with many development teams, OpenAI is responding with a...
The competition for the most effective AI coding assistant is intensifying. While Anthropic's Claude Code has recently gained favor with many development teams, OpenAI is responding with a substantial upgrade to its own tool, Codex.
The most significant change allows Codex to operate in the background on a Mac, controlling the cursor and applications independently. This enables the AI to perform tasks—like testing an app or iterating on frontend code—while a developer continues their primary work on the same machine. OpenAI positions this as an 'agentic' assistant handling auxiliary duties.
This move appears to directly counter a feature Anthropic announced last month, where Claude can remotely control a desktop. OpenAI's update also includes an in-app browser for executing commands on web applications, a feature aimed at frontend and game developers.
Beyond desktop control, the revamped Codex introduces several new capabilities. A 'memory' function helps the tool recall context from past sessions. It can now generate images for mockups and slide decks, and it integrates with over 110 third-party plugins, including CodeRabbit and GitLab Issues. These integrations allow Codex to perform organizational tasks, such as synthesizing a daily to-do list from Slack and Google Calendar.
OpenAI also announced a pay-as-you-go pricing option for its business and enterprise customers, providing more flexibility. The updates signal a clear strategic push to make Codex a multifaceted platform for corporate development workflows, marking another chapter in its ongoing contest with Anthropic for enterprise AI dominance.
Source: TechCrunch
Ready to Modernize Your Business?
Get your AI automation roadmap in minutes, not months.
Analyze Your Workflows →