Introducing OpenAI’s GPT-5.1 Codex Max: The Advanced Agentic Coding Model for Extended Productivity

OpenAI has launched GPT-5.1-Codex-Max, an advanced coding model designed for extended software development tasks. This upgrade boasts increased speed, intelligence, and token efficiency throughout the development cycle. Following Google’s introduction of its AI platform, Antigravity, the competition in AI-driven software development intensifies. Built on the GPT-5.1 architecture, Codex-Max was trained on real software engineering scenarios, leading to superior performance in coding evaluations. Notably, it operates effectively on Windows and excels as a team player with the Codex command-line tool. Accessible to ChatGPT Plus, Pro, Business, Edu, and Enterprise users, it outperforms its predecessor, achieving 79.9% accuracy on SWE-Lancer tests. Furthermore, efficiency is enhanced with reduced “thinking tokens,” translating to lower costs for developers. OpenAI’s commitment to continual improvement is evident, setting Codex-Max as the new benchmark for coding solutions in the AI landscape.

Source link