The past few months have surfaced an OpenAI versus Anthropic landscape. The Sam Altman-led company is strengthening its agentic coding and development system, Codex, with a number of updates that allow it to use your computer, generate images, and operate in the background. The announcement comes close on the heels of the success of Claude Code.
Codex will be able to operate directly within a user’s computing environment. The system can interact with applications on a machine by effectively “seeing” the screen and performing actions such as clicking and typing via its own cursor. This approach allows multiple AI agents to run tasks in parallel without interrupting the user’s own activity, potentially reducing friction in areas where traditional APIs are absent or limited. “For developers, this is helpful for iterating on frontend changes, testing apps, or working in applications that do not expose an API,” said OpenAI.
The feature will begin rolling out to Codex desktop app users signed in with ChatGPT and will initially be limited to macOS. OpenAI did not specify a timeline for when this capability will expand to other operating systems.
Codex will now be able to use GPT-Image-1.5 to generate and iterate on images. There are new plugins for tools such as GitLab, Atlassian Rovo, and Microsoft Suite, alongside native web browsing through an in-app browser, “where you can comment directly on pages to provide precise instructions to the agent”. The addition of an in-app browser extends Codex’s operational scope. Developers can annotate live web pages with instructions, enabling more precise feedback loops when working on interfaces, games, or web applications.
The update also significantly expands Codex’s integration ecosystem. More than 90 new plugins have been introduced, combining application integrations, skills, and MCP servers. These include connections to widely used developer and productivity platforms such as Jira, GitLab, CircleCI, and Microsoft Office. The practical implication is that Codex can now gather context and take action across a broader set of tools, reducing the need for manual switching between environments.
Within the application itself, the update introduces deeper support for developer workflows. Codex can now assist with pull request reviews, manage multiple terminal sessions, and connect to remote development environments via SSH, albeit in an early-stage release. File handling has also been enhanced, with richer previews for documents such as PDFs and spreadsheets, alongside a summary pane that surfaces agent plans, sources, and outputs.
