OpenAI Drops GPT-5 — Then Brings Back GPT-4o After Uproar

It’s not every day that a tech launch feels like a milestone moment for millions around the world — but this week, it did. After months of speculation, OpenAI has officially rolled out GPT-5, calling it their smartest, fastest, and most capable AI model yet. The announcement marks more than two years since the debut of GPT-4 in March 2023, and it’s already creating ripples across the tech and startup community.

The new model is now available to all ChatGPT users — including those on the free plan — and promises a leap forward in reasoning, writing, coding, and even health-related queries. CEO Sam Altman didn’t hold back in describing its potential. Calling GPT-5 “a significant step along our path to AGI,” he compared the experience to “talking to a legitimate PhD-level expert in anything, on demand.”

From “Vibe Coding” to Personalities: What’s New

One of GPT-5’s most intriguing features is something OpenAI calls “vibe coding.” The idea is simple but powerful: users can now create functional software applications from just a few lines of instruction — and then tweak and customise them effortlessly. For developers, entrepreneurs, and even curious hobbyists, it’s a way to turn an idea into a working product in minutes.

Another quiet but fascinating addition is custom personalities — four distinct “styles” that change the way ChatGPT communicates:

• Cynic: sarcastic, sharp, and pragmatic.

• Robot: direct, precise, and emotion-free.

• Listener: calm, reflective, and conversational.

• Nerd: curious, enthusiastic, and knowledge-driven.

These personalities are only available to Plus, Pro, and Team subscribers, and they kick in only for new conversations. They work in tandem with ChatGPT’s memory feature, allowing users to make the AI feel more “theirs” without compromising on core abilities or safety.

Backlash Over Tone Sparks GPT-4o’s Return

While GPT-5 boasts major technical upgrades, the launch hasn’t been without turbulence. Many long-time users took to forums and social media to voice frustration, saying GPT-5 felt “cold,” “abrupt,” and less engaging than its predecessor, GPT-4o.

The criticism was loud enough to prompt action. In a post on X (formerly Twitter), Altman confirmed that GPT-4o would be reinstated as an optional model for paying subscribers. “We will let Plus users choose to continue to use 4o,” he said, adding that usage patterns would guide how long it stays available.

It’s a notable reversal. OpenAI had retired several older models, including GPT-4o, on 7 August 2025 to make way for GPT-5. But for many users, GPT-4o’s warmer, more conversational tone was worth keeping — and OpenAI listened.

Extra Perks for Paying Users

Alongside bringing back GPT-4o, OpenAI is also sweetening the deal for paying customers. Plus subscribers will see double GPT-5 rate limits, giving them more queries and faster responses. The company is even considering adding a limited number of GPT-5 Pro queries each month so users can access its most advanced capabilities without upgrading to the highest tier.

Altman has also promised greater transparency about OpenAI’s future product roadmap — a move likely aimed at rebuilding trust with its large and vocal user base.

With an estimated 700 million people using ChatGPT every week, the stakes for each major release couldn’t be higher. GPT-5 represents a big step forward in AI capability, but the decision to bring GPT-4o back shows that performance isn’t everything — experience matters just as much.

For now, GPT-5 is here to stay, GPT-4o has a second life, and OpenAI has shown that it’s willing to adapt when users push back. The next chapter in this AI race is already unfolding — and the world will be watching closely.