Vercel

OpenAI Releases GPT 5.3 Codex, a Faster, Unified Coding Model


Executive Summary

A new model, GPT 5.3 Codex, is now available on the AI Gateway platform. This release combines the coding strengths of GPT-5.2-Codex with the reasoning capabilities of GPT-5.2 into a single, unified model that is 25% faster and more token-efficient. It is specifically designed for complex, agent-like tasks across the entire software development lifecycle, from research and debugging to deployment and documentation.

Key Takeaways

* Unified Model: Merges the distinct coding and reasoning capabilities of its predecessors into one model.

* Performance Improvements: Delivers a 25% speed increase and is more token-efficient.

* Agentic Capabilities: Built for long-running, multi-step tasks including research, tool use, debugging, and deployment.

* Interactive Control: Allows users to steer the model mid-task without losing the existing context.

* Enhanced Web Development: Better understands underspecified prompts and defaults to generating more functional, production-ready code.

* Availability: Available now on AI Gateway using the model identifier `openai/gpt-5.3-codex`.

Strategic Importance

This launch streamlines the development workflow by providing a single, more powerful tool for complex software engineering tasks. It reinforces the company's focus on creating sophisticated, agentic AI assistants for professional developers, increasing the platform's utility for enterprise use cases.

Original article