OpenAI rolls out ChatGPT-5 free for all users

SAN FRANCISCO: OpenAI has launched ChatGPT-5, its most advanced version yet — and it’s available to everyone for free as the global race to dominate the technology heats up. 

The new model, which already has millions of users each week, is being praised for its smarter, faster responses and wide range of skills.

The company’s Co-founder and chief executive Sam Altman said that it feels like talking to a real expert, describing the latest model as “clearly a model that is generally intelligent.”

While it's not quite thinking like a human yet, he believes it’s a big step closer. 

The release comes as tech giants race to lead the fast-moving world of artificial intelligence.

ChatGPT-5 is now rolling out free of charge to all users, with the AI tool currently being used by nearly 700 million people each week, OpenAI said during a media briefing.

However, he stressed that more progress is needed to achieve artificial general intelligence (AGI) — the kind of AI that can think like a human being.

“This is not a model that continuously learns from new information it encounters during use — which, to me, feels like it should be part of AGI,” Altman explained.

“But the level of capability here is a huge improvement.”

Industry experts have hailed the beginning of a new AI era — one in which intelligent machines change the way people live and work.

“As the pace of AI progress accelerates, developing superintelligence is coming into sight,” wrote Meta chief executive Mark Zuckerberg in a recent internal memo.

“I believe this will be the beginning of a new era for humanity.”

Altman said there were “orders of magnitude more gains” still to come on the road to AGI.

“Obviously... you have to invest in compute power at an eye-watering rate to get there, but we intend to keep doing it,” he said.

Tech giants such as Amazon, Google, Meta, Microsoft, and Elon Musk’s xAI have been investing billions of dollars into AI research and development since the launch of the first ChatGPT model in late 2022.

Earlier this year, Chinese start-up DeepSeek surprised the AI sector by releasing a powerful model that delivers high performance using more affordable chips.

‘PhD-level expert’

With fierce global competition in the AI space, Altman claimed that ChatGPT-5 is now leading the field in areas like coding, writing, healthcare and more.

“GPT-3 felt like talking to a secondary school student — ask a question, you might get a correct answer or something completely off,” Altman said.

“GPT-4 was more like speaking to a university student. GPT-5 is the first time it really feels like you’re chatting with a PhD-level expert on any topic.”

Altman believes that “vibe-coding” — the ability to build software through natural conversation — will become a defining feature of the ChatGPT-5 era.

British AI expert Simon Willison, who had early access to the model, wrote in a blog post: “My verdict: it’s just good at stuff. It doesn’t feel like a dramatic leap forward, but it rarely messes up — and often impresses.”

However, Elon Musk claimed on X (formerly Twitter) that his own AI model, Grok 4 Heavy, was “smarter” than ChatGPT-5.

Honest AI?

According to OpenAI’s safety research lead Alex Beutel, ChatGPT-5 has been trained to be trustworthy and avoid aiding any harmful tasks.

“We built evaluations to measure how often it may deceive, and trained the model to be honest,” Beutel said.

The model is designed to produce “safe completions” — responses that avoid sharing sensitive or potentially dangerous information, he added.

Also this week, OpenAI launched two new AI models that are freely available to download and customise, in a move aimed at challenging similar offerings from competitors.

The release of these “open-weight language models” comes as the company faces growing calls to be more transparent, in line with its original mission as a non-profit organisation.