That’s because OpenAI just officially launched Codex, a stunning new AI coding companion. We’re launching today on ChatGPT Pro, Enterprise and Team subscribers. They’re getting hands-on experience with one of the most sophisticated AI coding agents our company has developed! Codex is based on the codex-1 model. Based on OpenAI’s o3 AI reasoning model, this newer, more fine-tuned version is optimized specifically for software engineering tasks.
According to Codex’s creators, it’s productivity with a magic touch—users will simply be able to make Codex handle lots of software engineering tasks in parallel. It works quietly in the background. This enables an uninterrupted user experience to access your computer and browser while Codex runs commands in the background. This flexibility is designed to make developers and engineers’ workflows more efficient.
Codex OpenAI has promised users “generous access” to Codex at first. In the weeks ahead, we will be introducing rate limits for this use case. This change will make sure that all subscribers can use the powerful tool fairly and effectively. Through its integration with GitHub, Codex preloads users’ code repositories into its environment. This feature makes it a lot more usable and fits neatly into any current development process.
OpenAI’s API already has Codex baked right into it.
Pricing
Pricing is $1.50 per 1 million input tokens and $6 per 1 million output tokens. Through this new pricing model, we want to make powerful new coding capabilities accessible and affordable for developers of all skill levels.
Alexander Embiricos, OpenAI’s Product Lead, Safety was paramount during the development of Codex. As many of the safety measures created for OpenAI’s o3 model are relevant to Codex, he explained. Safety should be top priority as the tool enters a burgeoning marketplace. For instance, Anthropic and Google have recently released their own AI coding solutions, Claude Code and Gemini Code Assist respectively.
The launch of Codex is part of OpenAI’s broader strategy to continuously enhance ChatGPT by integrating additional products that cater to diverse user needs. The platform gives users more control by allowing them to self-track progress more easily. We show all of the other tasks Codex has been given, encouraging a collaborative and open coding experience.
AI coding tools have taken the world by storm in the past few years. At the same time, amazing new platforms like Cursor are already generating upwards of $300 million in annualized revenue. This trend is no surprise as there is an increasing demand for automated solutions that help developers write better code faster.
Like any other cutting-edge technology, OpenAI understands that AI coding tools pose a number of risks. The firm admitted to issues of “malcontent software.” They are warning against the temptation to take our eyes off the ball as users start to bring Codex into their workflows.
Leave a Reply