OpenAI has launched GPT-5-Codex, a new model designed to enhance developer productivity and automate coding tasks. This release is specifically targeted at software engineers, data scientists, and developers who rely on AI-powered coding tools. The model is now accessible to users across all Codex interfaces, including integrated development environments and web-based code editors.
GPT-5-Codex introduces agentic coding abilities, enabling it to undertake extended development sessions, maintain context over large codebases, and execute multi-step programming tasks with minimal human intervention. Compared with previous Codex versions, GPT-5-Codex offers more robust code understanding, supports a larger range of programming languages, and can handle more intricate project requirements.
The model leverages advancements from the GPT-5 foundation, including a larger context window and improved reasoning capabilities. These enhancements make it suitable for both collaborative pair-programming and fully autonomous code generation. Early user feedback highlights its ability to reduce errors and accelerate project timelines, while industry analysts note its potential to shift workflows in software development.
GPT-5-Codex is here: a version of GPT-5 better at agentic coding.
— Sam Altman (@sama) September 15, 2025
It is faster, smarter, and has new capabilities. Let us know what you think!
The team has been absolutely cooking, very fun to watch.
The new model is now rolling out in Codex products — which can be accessed via a terminal, IDE, GitHub, or ChatGPT — to all ChatGPT Plus, Pro, Business, Edu, and Enterprise users.
OpenAI, known for its leadership in large language models, continues to develop AI tools specifically tailored to real-world engineering problems. This release reflects OpenAI’s ongoing focus on practical applications of generative AI in technical fields, aiming to provide developers with more dependable and adaptable coding assistants.
