• Source:JND
Add as a preferred

OpenAI’s latest frontier AI model, GPT-5, launched Thursday as a unification of the GPT-series and o-series. The idea was to create a single, all-in-one model so users wouldn’t need to constantly switch via the model picker dropdown. Alongside the launch, OpenAI retired older models—including the widely loved GPT-4o.

That decision hasn’t gone down well. Across social media and Reddit, many ChatGPT users are now complaining that GPT-5 feels like a downgrade.

Why Users Want GPT-4o Back

On Reddit, complaints about GPT-5 often centre on shorter, less detailed responses and a loss of “personality.”

In an AMA with OpenAI CEO Sam Altman, one user summed it up:

“I’ve been using GPT-4o for a long time and have built a very specific dynamic and workflow with it. I know GPT-5 is designed to be stronger for complex reasoning, coding, and professional tasks, but not all of us need a pro coding model. Some of us rely on 4o for creative collaboration, emotional nuance, roleplay, and other long-form, high-context interactions.”

Altman acknowledged the criticism and promised fixes.

ALSO READ: India Set To Surpass US As Top ChatGPT Market, Says Sam Altman During GPT-5 Launch

Altman’s Response: Rate Limit Boosts and Model Access

A few hours later, Altman posted on X (formerly Twitter) outlining changes coming within days:

  • Double GPT-5 rate limits for Plus subscribers (to address complaints about hitting limits within an hour)
  • Continued access to GPT-4o for Plus subscribers—though OpenAI will monitor usage before deciding how long it remains available

Altman also revealed a technical reason GPT-5 may have felt “dumber.” The model’s autoswitcher—a real-time router that shifts between capabilities (reasoning, conversational, creative)—was temporarily “out of commission.” Without it, GPT-5 couldn’t adjust properly to different queries.

OpenAI is now fixing the routing system and will also start showing users exactly which model is generating a given response.

The Business Reality: OpenAI Is Running at a Loss

Speaking to CNBC after the GPT-5 launch, Altman admitted OpenAI is currently operating at a loss, despite making billions in revenue.

“As long as we’re on this very distinct curve of the model getting better and better, I think the rational thing to do is to just be willing to run the loss for quite a while,” he said.

Last year, OpenAI generated around $3.7 billion in revenue but spent $5 billion, largely on infrastructure costs—both for in-house servers and rented cloud capacity.

The retirement of older models like GPT-4o likely ties back to these costs. Maintaining multiple large models is expensive, but dropping them risks alienating loyal users.

ALSO READ: Take-Two CEO Says GTA 6 Price Will Reflect Value, 'Our Goal Always Is To Deliver....'

The Catch-22

Each of OpenAI’s older models was trained with different strengths—GPT-4o excelled at creative, emotionally nuanced, and conversational tasks, while GPT-5 focuses more on reasoning and professional-grade work.

The company now faces a tricky balancing act:

  • Keep older models alive and continue burning cash
  • Or force users to adapt to GPT-5 and risk long-term dissatisfaction

For now, GPT-4o survives—at least for Plus subscribers—but its long-term availability will depend on how many people use it.

If you want, I can also make a punchier headline list for this story so it’s more clickable in search results. That way, you’ve got options to A/B test.