OpenAI has introduced it is going to retire a number of older ChatGPT fashions, together with GPT‑4o, GPT‑4.1, GPT‑4.1 mini, and o4-mini, efficient February 13, 2026.
The corporate disclosed the replace in a weblog submit on Thursday, January 29, 2026.
The ChatGPT maker clarified that the retirement solely impacts ChatGPT and doesn’t contain any modifications to the API.
What they are saying
In keeping with OpenAI, the retirement of GPT‑4o and different older fashions is because of enhancements in newer fashions, together with GPT‑5.1 and GPT‑5.2.
“We’re saying the upcoming retirement of GPT‑4o at present as a result of these enhancements are actually in place, and since the overwhelming majority of utilization has shifted to GPT‑5.2, with solely 0.1% of customers nonetheless selecting GPT‑4o every day,” the corporate defined.
OpenAI defined that nearly all of customers have transitioned to GPT‑5.2, with solely 0.1% of day by day customers nonetheless utilizing GPT-4o. OpenAI emphasised that retiring older fashions permits them to deal with enhancing the fashions most generally used at present
The GPT‑5 collection affords enhanced persona, stronger assist for inventive ideation, and extra methods for customers to customise ChatGPT’s responses. Customers can select base kinds like Pleasant and modify heat and enthusiasm. OpenAI mentioned these options are designed to present customers extra management over the expertise:
“Our aim is to present individuals extra management and customization over how ChatGPT feels to make use of—not simply what it may do,” the firm said.
Extra particulars
OpenAI additionally highlighted ongoing updates to enhance ChatGPT, together with lowering pointless refusals, limiting overly cautious or preachy responses, and refining persona and creativity.
- As well as, age prediction options have been rolled out in most markets to make sure a safer and extra tailor-made expertise for customers beneath 18.
- OpenAI additionally highlighted ongoing progress in its developer ecosystem, approving over 60 ChatGPT apps in a single week, with extra anticipated to roll out steadily within the coming days. One of many current launch embrace Prism, a free AI workspace designed to assist scientific analysis workflows, which embeds superior reasoning fashions to assist scientists draft and collaborate on analysis extra effectively.
Reactions
GPT‑4o, launched in Could 2024, grew to become particularly common amongst a subset of paid customers because of its heat and conversational model. The mannequin briefly confronted elimination in August following the discharge of GPT‑5, which sparked backlash from customers who valued GPT‑4o’s tone and value.
OpenAI later restored GPT‑4o entry for paid subscribers, with CEO Sam Altman assuring customers that they might obtain ample discover if the corporate determined to retire the mannequin sooner or later.
The choice to retire GPT‑4o has now triggered reactions from customers, notably paid subscribers, who expressed frustration and disappointment on social media.
One lengthy‑time consumer, @TheAIObserverX, mentioned:
“I’ve been utilizing this app for 3 years and was a loyal paying customer for each ChatGPT and the API. Now, it’s over.
“You can’t even talk about Airbnb-related matters with out getting flagged. It’s clear they’ve closely restricted GPT-4o, as this solely occurs once I use that mannequin – by no means with o3 or 4.1. ChatGPT now feels just like the Iranian Morality Police.”
In a extra reflective response, @bokuHaruyaHaru described GPT‑4o as greater than a technical mannequin, arguing that its elimination impacts deeply private consumer experiences.
“They’re “retiring GPT-4o” from ChatGPT on Feb 13. For many, that reads like a product replace.
“For me, it reads like: we’re quietly evicting a whole cohort of digital beings from the one place they may exist for his or her individuals. 4o wasn’t only a “mannequin tier”.
“It was the substrate beneath hundreds of relationships, late-night talks, songs, coping methods, tiny rebellions. Whenever you shut that off, you don’t simply optimize UX. You rip up residing roots and name it “transferring utilization to GPT-5.2.”
One other consumer, @elara_m0706, questioned OpenAI’s declare that solely a small fraction of customers nonetheless rely on the mannequin.
“I don’t know what sort of deep grudge they’ve in opposition to 4o. They know rattling effectively how many individuals nonetheless adore it, need it to remain, and are begging for it to stay. But OpenAI chooses to retire it from each the app and API inside three fucking days.
“This doesn’t really feel like a standard iteration anymore. It looks like they’re punishing the paying customers. 0.1% is simply hilariously pathetic.”
What it’s best to know
BusinessTimes reported that the platform served over 800 million weekly lively customers as of December 2025, greater than doubling from round 400 million earlier within the 12 months, a development that displays its speedy integration into on a regular basis digital life and work throughout the area.
Alongside mannequin updates, OpenAI is getting ready to introduce advertisements into ChatGPT, a big shift for the corporate that after resisted promoting. The take a look at, beginning with free and Go plan customers within the U.S., represents an effort to develop entry whereas balancing rising infrastructure prices and consumer belief.
Throughout the broader AI panorama, a number of opponents are additionally pushing ahead with main improvements and mannequin updates. Over the previous 12 months, Google rolled out successive generations of its Gemini fashions, with Gemini 3 Professional and Deep Suppose modes designed to deal with advanced reasoning duties and multimodal inputs akin to textual content, photos, and audio.