The memory feature of Chatgpt is now available to free users; Codex expansion for paid customers | Mint

Openai has begun to roll an upgraded memory feature for users of the free version of Chatgpt, a tool that was exclusive to those on the Pro Plan earlier. The enhancement allows AI Chatbot to provide more analogous reactions by recalling user preferences and recent interactions.

The announced in April and gradually being extended to more users, the updated memory system enables Chatgpt to provide better support in activities such as writing, learning and general advice by referring to recent interactions. Openai confirmed development on X, stating, “We are starting rolling a light version of memory reforms for free users. In addition to the existing saved memories, Chatgpt now refers to your recent conversations to provide more personal reactions.”

The users can test how much the chatbot is remembered only by motivating the chatbot to “describe me based on our chat”. Although the model cannot miss the conversation from several months in advance, it maintains information from more recent sessions.

For those who prefer to use a memory function, OpenIAI offers options to disable it. Can be found under setting Settings> Security> MemoryWhere users can fully switch the memory or review and remove the already stored information.

In parallel, Openai has also expanded access to its Codex coding accessory. Initially only enterprise, or pro, or pro available to the pro subscribers, the tool is now accessible to the Chatgpt Plus users. Codex is designed to support developers with writing, debugging and test codes.

This update has a significant new facility of Codex to reach the Internet during programming works. With enabling this function, the supporting package can install, the staging can interact with the environment, and obtain the resources required to run specific tests. However, internet access is disabled by default and must be activated manually. Users can also control which domains can reach Codex and allow HTTP methods to be used.

Openai said it is monitoring the system for any possible misuse, including the so -called “early injection attacks”, to ensure that the coding assistant safely and as intends.