OpenAI brings Custom GPTs, memory, vision and more to free ChatGPT users

ChatGPT maker OpenAI says users on the free tier can now use GPT-4o features like data analysis, memory, ask questions about images and more.

ChatGPT | ChatGPT new features | ChatGPT GPT-4o

ChatGPT can now analyse charts and images and remember things you tell it to. (Express Photo)

OpenAI has announced that free ChatGPT users can now access custom GPTs created by others, ask questions about images, ask the chatbot to analyse charts and use features introduced earlier this month with GPT-4o.

Compared to GPT-4, OpenAI says GPT-4o is a multimodal large language model (LLM) much faster and improves existing text, vision and audio capabilities. The company’s CTO Mira Murati had said that the model will be available for free users in the coming days and it looks like OpenAI is rolling out the new features to everyone.

This means users can use GPT-4o features like data analytics, vision, memory and upload files for analysis. The new model also reduces the latency when using voice to interact with the AI chatbot.

Also Read | AI can enable voters to make informed decisions, mitigate polarisation during elections: Study

GPT-4o can also understand things like tone and background noises, and even insert emotions in its voice. It also offers multilingual support.

While ChatGPT users on the free tier can use the new GPT-4o-powered functionalities, OpenAI says paid subscribers will have “up to five times the capacity limits” of free users. Once users exhaust their GPT-4o limits, ChatGPT will automatically switch to GPT-3.5 to continue the conversation. OpenAI says GPT-4o is safer since it includes “filtered training data, and refined model behaviour post training.”

OpenAI recently announced that it has formed a new ‘Safety and Security Committee’ led by CEO Sam Altman and said that the company is now working on a new AI model, which is rumoured to be GPT-5.