OpenAI has released a major update to its Codex desktop app, expanding the tool beyond code generation into a broader software development and productivity environment used by more than 3 million developers each week.
The update introduces computer control, an in-app browser, image generation, memory, and expanded automation. The release reflects a shift in how AI tools integrate into day-to-day development workflows rather than operate as isolated assistants.
Codex for (almost) everything.
— OpenAI (@OpenAI) April 16, 2026
It can now use apps on your Mac, connect to more of your tools, create images, learn from previous actions, remember how you like to work, and take on ongoing and repeatable tasks. pic.twitter.com/UEEsYBDYfo
Codex moves beyond code into full workflows
Codex started as a coding assistant. The new version extends across the full software lifecycle, from writing and testing code to reviewing changes and coordinating tasks.
The app now supports multiple terminal tabs, GitHub pull request review comments, and connections to remote development environments through SSH in alpha. Developers can open files directly inside the interface, with previews for documents such as PDFs, spreadsheets, and slides.
A new summary pane tracks agent plans, sources, and outputs. This structure keeps context visible across tasks and reduces the need to switch between tools.
OpenAI described the goal as enabling developers to move across workflows in one environment rather than split work across separate systems.
Computer use allows direct control of applications
One of the most notable additions is computer use. Codex can now interact with applications on a user’s computer by seeing the screen, moving a cursor, and typing commands.
This feature runs in the background. Users can continue their own work while the agent executes tasks in parallel. OpenAI noted that multiple agents can operate at the same time without interfering with other applications.
The functionality targets workflows that lack APIs. Developers can test apps, adjust frontend elements, or perform repetitive tasks directly through the interface.
Computer use is currently available on macOS, with expansion to other regions planned.
Built-in browser improves frontend iteration
The updated Codex includes an in-app browser. Users can interact with web pages and give precise instructions by clicking on elements.
This approach reduces ambiguity in frontend work. Instead of describing changes in text, developers can select specific elements and request edits directly.
OpenAI stated that the browser will expand over time. The long-term plan includes broader control beyond local web applications.
Image generation enters developer workflows
Codex now integrates image generation through gpt-image-1.5. Developers can create visuals such as mockups, product concepts, and game assets within the same workflow as code.
This addition removes the need to switch tools for design tasks. It also allows agents to combine screenshots, code, and generated images to verify outputs.
The feature aligns with use cases in frontend development and prototyping.
Plugins extend access across tools
OpenAI introduced more than 90 new plugins to expand Codex capabilities. These integrations connect the app to widely used developer and enterprise tools.
Examples include Atlassian Rovo for Jira management, CircleCI for continuous integration, GitLab Issues, the Microsoft Suite, and Neon by Databricks. Plugins combine app integrations, skills, and model context protocol servers to allow Codex to gather context and take action across systems.
OpenAI stated that plugins are curated before release. This approach addresses concerns around security and reliability.
Automation and memory support long-term tasks
The update expands automation features. Codex can now reuse existing conversation threads and continue tasks over extended periods.
OpenAI explained that the system can “wake up automatically to continue on a long-term task, potentially across days or weeks.” Teams use these automations for managing pull requests, tracking conversations, and following up on tasks across tools such as Slack, Gmail, and Notion.
The app also introduces a memory feature in preview. Codex can store user preferences, corrections, and previously gathered context. This allows future tasks to complete faster and with fewer instructions.
OpenAI stated:
“This helps future tasks complete faster and to a level of quality previously only possible through extensive custom instructions.”
Proactive suggestions guide daily work
Codex now suggests actions based on project context. The system can identify pending tasks, such as unresolved comments or incomplete workflows, and present a prioritized list.
The feature draws from connected tools and stored context. It can surface relevant information from documents, communication platforms, and codebases.
OpenAI described this as a way to help users resume work without manual tracking.
Gradual rollout with broader vision ahead
The update is rolling out to Codex desktop users who sign in with ChatGPT. Some features, including memory and personalization, will reach enterprise and regional users later.
The release offers insight into OpenAI’s broader direction. During a briefing, Codex lead Thibault Sottiaux said:
“We’re building the super app out in the open. This release is about developers. In the future, we will broaden it up to a wider audience.”
The company continues to position Codex as a central interface for development tasks. The integration of coding, automation, and system control reflects a move toward unified AI workspaces.
Anastasiia G
Disclaimer: All materials on this site are for informational purposes only. None of the material should be interpreted as investment advice. Please note that, despite the nature of much of the material created and hosted on this website, HODL FM operates as a media and informational platform, not a provider of financial advisory services. The opinions of authors and other contributors are their own and should not be taken as financial advice. If you require advice, HODL FM strongly recommends contacting a qualified industry professional.
