OpenAI has announced a sweeping upgrade to its Codex platform, introducing GPT-5-Codex—a specialized variant of GPT-5 designed to deliver superior coding automation, smarter real-time collaboration, and robust independent task execution across development environments. The update unites Codex tools under a single, streamlined product, promising an AI teammate that blends seamlessly into developers’ workflows—whether at the terminal, in the IDE, on the web, or on mobile devices.

I needed to update a codebase owned by another team for a feature release. [...] With Codex, I offloaded the refactoring and test generation while focusing on other priorities. It produced high-quality, fully tested code that I could quickly hand back — keeping the feature on schedule without adding risk.

Tres Wong-Godfrey, Tech Lead at Cisco Meraki

OpenAI Unveils GPT-5-Codex

OpenAI’s rollout of GPT-5-Codex dramatically raises the bar for AI-assisted coding. The new model is engineered for real-world software engineering, improving reliability in both interactive sessions and fully autonomous coding sprints. GPT-5-Codex boasts enhanced capabilities, including independent project builds, sophisticated feature additions and debugging, end-to-end refactoring, and detailed code review, all fine-tuned for agentic workflows. Developers can expect higher code quality and fewer bugs shipped as the model’s code review abilities catch issues early—moving well beyond basic completions into the realm of a trustworthy engineering partner.

Unified Experience Across CLI, IDE, and Cloud

The upgrade brings Codex into a single, robust experience linked by your ChatGPT account, erasing the friction between local and cloud-based development. Codex now works smoothly across command-line, IDE extensions (for VS Code, Cursor, and derivatives), browser-based tools, and even the ChatGPT iOS app. Developers can start a project in their terminal and seamlessly hand it off to the cloud, or collaborate in real-time in their favorite IDE—with full context preserved throughout the process.

Next-Generation Code Review & Security

OpenAI has fortified Codex’s code review: GPT-5-Codex now automatically analyzes pull requests, inspects dependencies, runs code/tests, and flags critical flaws with high accuracy. Benchmarks show a leap from 33.9% to 51.3% in code refactoring performance and a major reduction in irrelevant review comments, highlighting only the most impactful issues and saving developers’ time.

Security is at the forefront: Codex operates in a sandboxed environment with network access disabled by default, minimizing risks from harmful code or prompt injection. Customizable permission settings and audit trails empower teams to balance productivity with safety, while comprehensive documentation and logs ensure accountability.

Performance That Adapts to Your Workflow

Codex dynamically shifts how much time it spends “thinking” based on task complexity—delivering rapid responses for small fixes or investing hours in complex refactors and bug hunts. During testing, GPT-5-Codex successfully ran uninterrupted for over seven hours on intricate engineering challenges, iterating, debugging, and delivering robust solutions autonomously.

Availability, Access, and Developer-Friendly Pricing

GPT-5-Codex is now standard across all ChatGPT Plus, Pro, Business, Edu, and Enterprise plans, and access via API is slated for release later this year. Business and Enterprise customers receive scalable credit pools for collaborative development, with flexible options to match usage patterns. Developers can access Codex across platforms, eliminating barriers for team productivity and innovation. For more details, click here.