OpenAI has officially launched Codex, a new AI-powered coding agent designed to streamline software development by assisting with tasks like writing, testing, and debugging code — all autonomously.
Built on the codex-1 model, which leverages the “reasoning” variant of GPT-4 (o3), Codex is engineered to produce clean, functional code and can independently iterate until programs run without errors. With a context window of up to 192,000 tokens, it can manage large-scale projects and complex codebases.
Key Features of Codex:
- Runs entirely in the cloud, reducing browser load and allowing for seamless use on desktops and smartphones.
- Supports multi-tasking, integration with GitHub repositories, and automatic generation of code change logs.
- Available now for Pro, Enterprise, and Team users, with Plus and Edu users gaining access in the near future.
- Will be accessible through the ChatGPT sidebar and a dedicated window.
In the coming weeks, OpenAI plans to roll out speed limits on the agent and introduce a credit-based system for extended use.
OpenAI has also released Codex-Mini, a more lightweight version powered by the o4-mini model, designed for developers seeking lower-cost integration. Pricing for Codex-Mini is set at $1.5 per million input tokens and $6 per million output tokens, available via API.
