OpenAI releases GPT-5.1, boosting speed, intelligence, and developer features like 24-hour prompt caching and new tools.
OpenAI unveiled GPT-5.1 on November 13, 2025, presenting a major advancement in balancing model intelligence and responsiveness. This iteration dynamically adjusts thinking time based on task complexity to significantly enhance speed and token efficiency across diverse tasks such as coding and agentic reasoning. A novel 24-hour extended prompt caching feature allows the model to remember context longer, improving interaction continuity for developers. Moreover, GPT-5.1 introduces fresh tooling including ‘apply_patch’ and ‘shell’ commands that expand its functionality, along with a ‘no reasoning’ mode optimized for rapid responses on straightforward queries. These upgrades position GPT-5.1 as a versatile and efficient model tailored to practical developer needs and complex AI applications.[1]