OpenAI's Codex has received a major update. Engineer Tibo, who is responsible for Codex and ChatGPT, officially announced on a social platform that the highly anticipated GPT-5.6Sol model's million-Token context window feature is now available to all ChatGPT accounts.

Previously, this large context capability was only available through API key calls. With the removal of this restriction, users with ChatGPT Plus and Pro subscription plans can now manually configure it to enjoy the full context window of about 1.05 million Tokens in the Codex programming proxy tool.

image.png

For developers, this means Codex can now accommodate and process more code, tool outputs, and complete conversation records before automatically compressing historical information. Whether it's refactoring large codebases, deep debugging of complex projects, or long and intensive sessions, this feature brings great convenience and completely solves the pain point of previous context limitations affecting large project maintenance.

However, the development team also issued a friendly reminder. The current default context length is a carefully optimized compromise between performance and cost. Blindly setting the maximum limit may easily consume your weekly quota quickly and lead to unnecessary waste. Therefore, developers are advised to enable it according to specific business scenarios, allowing the large model to operate efficiently within the most suitable range.