If you’re using one of the best laptops or best smartphones, chances are, you’re also dabbling in AI. If so, a free ChatGPT update might be worth looking into – OpenAI has started rolling out a new default model, and it’s aimed at making answers more accurate, less waffle-heavy, and more useful.
The upgrade swaps in GPT-5.5 Instant as the new standard model inside ChatGPT, replacing the previous GPT-5.3 Instant. According to the company, it’s rolling out today to all users, including those on the free tier.
The headline feature is accuracy. OpenAI says that the new model produces far fewer hallucinations – particularly in high-stakes areas like medicine, law, and finance – and reduces hallucinated claims by more than 50 per cent in internal testing. It also apparently improves performance on things like image analysis, maths, and general reasoning tasks.
Responses are also designed to be tighter and more to the point, cutting down on overlong explanations, unnecessary formatting, and, thankfully, fewer random emojis.
There’s also a push towards smarter personalisation. GPT-5.5 Instant is better at selectively using context from past chats, files, and connected accounts like Gmail to tailor responses. That feature is rolling out first to Plus and Pro users on the web, with mobile support and wider availability expected in the coming weeks.
Another new addition is memory sources – a way to see what past context the model used to generate a response. It’s designed to give users more transparency and control, including the option to delete or tweak stored information if something’s out of date.
We’ve yet to try it out for ourselves, but hopefully the new ChatGPT update is a step in the right direction, and won’t break anything that already works. Yes, we’ve been burned before.
Read the full article here
