OpenAI Launches Codex: A Software Engineering Agent In ChatGPT

Codex can write codes, fix bugs, run tests, and manage multiple coding tasks at once. According to Greg Brockman, 2025 will be the year of agents.

OpenAI continues to make ChatGPT useful for developers.

A few days ago, they added support to connect GitHub repositories to deep research and ask questions based on your own code.

Today, the company launched a research preview of Codex in ChatGPT, its most capable AI coding agent yet. It can write code, fix bugs, run tests, and manage multiple coding tasks at once — all within a secure cloud environment.

Let’s break down the details of the Codex announcement.

• What is Codex?

• How Codex works

• Codex performance benchmarks

• Codex is not for vibe coding

• Codex can’t access the internet

• How to access Codex

• Final thoughts

What is Codex?

Codex is a software engineering agent that runs in the cloud and does tasks for you, like writing a new feature of fixing a bug.

You can run many tasks in parallel.