New Delhi: OpenAI has recently released its Codex application for Windows, bringing the desktop interface designed to help Windows users and developers manage multiple AI agents at once. This application is especially designed to help developers build, review, and ship software by using AI agents powered by OpenAI’s coding models.
The Codex applications act as the command centre where developers can manage AI agents and coding tasks. The release builds on the Codex’s existing availability via the Command Line Interface, IDE extensions, and cloud-based tools.
The Codex Windows application is designed for real development environments, and it includes native sandboxing and integrated workflows. This enables the developers to run coding agents without affecting their main system environments.
OpenAI has also mentioned that the application is designed to support the complex, long-running tasks that might span hours, days, or weeks. The users can switch between agent threads without losing context, review changes made by the agents, comment on diffs, or open files in their editor to make manual edits.
The application also carries over the session history and the configuration from the Codex CLI and IDE extensions, enabling users to continue working with existing projects without reconfiguration. The developers can run multiple AI agents across different projects at the same time.
The tool can also automate repetitive tasks that developers usually handle manually. The feature also enables agents to connect with other development tools and workflows through something OpenAI calls Skills.
The developers can also review and guide the work done by the AI agents without losing the context of the project. This also helps the developers stay in control of their code while still using automation to handle certain tasks.
The tool is already being used by many of the startups such as Ramp and Harvey, as well as well-known companies including NVIDIA, Rakuten, and Cisco. The Codex agent runs commands in PowerShell using the Windows native environment.
