With improved memory, ChatGPT can deliver more personalised responses to users based on past conversations.
OpenAI ChatGPT Memory Feature: This feature was previously only available to paid subscribers
Photo Credit: Reuters
OpenAI is now rolling out ChatGPT's memory improvements to those on the free tier of the platform. Announced on Tuesday, the San Francisco-based artificial intelligence (AI) firm announced that now all users of the chatbot will get personalised responses based on their previous conversations with ChatGPT. Notably, the free users are getting a lightweight version of the feature. Additionally, the company is also providing Codex, its coding-focused AI agent, full access to the Internet that can be activated during the setup phase.
In a post on X (formerly known as Twitter), the official handle of OpenAI announced the rollout of the new feature. The memory boost was first introduced in April to the ChatGPT Plus, Team, and Pro subscribers. However, now, it is also being expanded to those users who are on the free tier of the platform.
Before this memory improvement, ChatGPT could only save and remember specific information about users, which could be changed or deleted manually by visiting the Memory section in Settings. However, the current upgrade will also allow the AI chatbot to gain information and context from past conversations with the user.
This means whenever users start a new session, users will not have to repeat old information in a detailed prompt. Instead, they can start a normal conversation, and ChatGPT will already know the context. Users can also remind the chatbot about a particular conversation using a prompt. For instance, they can type, “Based on our conversation about the quarterly performance report, tell me what the goals should be for the next quarter,” and it will understand the context.
Notably, OpenAI highlighted that the free users will get a lightweight version of the feature, instead of what the paid subscribers are getting. However, the company did not mention what the changes would be.
Separately, the AI firm is also upgrading its coding agent Codex. In a post on X, OpenAI CEO Sam Altman announced that Codex will now have full access to the Internet, which can be activated during the setup phase. This setting will remain turned off by default.
In its support page, OpenAI has highlighted that enabling Internet connectivity for Codex could expose users and organisations to security risks. The company mentions risks such as prompt injection, exfiltration of code, and inclusion of malware. These risks occur as the AI agent navigates across the web and ends up retrieving or processing untrusted or harmful content.
Altman also mentioned that Codex is now also being made available to the ChatGPT Plus tier. Earlier, the AI agent was only available to Pro subscribers.
Catch the latest from the Consumer Electronics Show on Gadgets 360, at our CES 2026 hub.