Listen to this Post
A Familiar Face Disappears from ChatGPT
From time to time, OpenAI reshapes ChatGPT’s model lineup, quietly phasing out older systems as newer ones take center stage. This time, the change is far more emotional than technical. On February 13, OpenAI will officially retire several well-known models from ChatGPT’s model picker, including GPT-4o — a model many users didn’t just use, but bonded with. The move signals another major shift in how OpenAI wants people to experience AI conversations, even as it insists the vast majority of users have already moved on.
OpenAI’s Announcement and Model Retirements
OpenAI has confirmed that multiple models will be removed from ChatGPT’s interface in mid-February, with GPT-4o leading the list. The decision echoes a controversial moment from last August, when GPT-4o was abruptly removed following the launch of GPT-5. That initial removal triggered immediate backlash across the community, as users complained that GPT-5 felt colder, more mechanical, and less empathetic than GPT-4o. Many described GPT-4o as warmer, friendlier, and more “human” in tone, making it their preferred conversational partner.
The backlash was strong enough that OpenAI reversed course and restored GPT-4o, with CEO Sam Altman publicly promising advance notice before any future deprecation. He also acknowledged that users form emotional attachments to language models, an issue that sparked broader ethical debates about AI companionship and personalization. Since then, OpenAI has rolled out expanded personality customization, allowing users to choose from seven predefined personalities or craft their own custom instructions.
Despite these changes, OpenAI now says GPT-4o accounts for only about 0.1% of daily user interactions, arguing that its continued presence is no longer justified. Alongside GPT-4o, the company will also retire GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini from ChatGPT’s model picker. Importantly, OpenAI emphasized that these changes affect only ChatGPT’s interface; the models will remain available via the API for developers, at least for now. The company has not announced a firm timeline for API deprecation, leaving the door open for continued backend use even as everyday users lose access.
What Undercode Say:
This retirement is less about usage statistics and more about control. OpenAI is clearly signaling that ChatGPT’s future is streamlined, personality-tunable, and centered around fewer flagship models rather than a crowded picker of legacy options. While the company frames the move as a response to user behavior, the 0.1% figure feels more like justification than explanation. GPT-4o’s influence extended far beyond raw usage numbers — it shaped expectations around what conversational AI should feel like.
The emotional backlash surrounding GPT-4o exposed an uncomfortable truth for AI companies: users don’t see models as interchangeable utilities. They perceive them as distinct entities with tone, temperament, and personality. OpenAI’s answer to this problem wasn’t to preserve older models indefinitely, but to abstract “personality” away from the model itself. By offering customizable personalities layered on top of GPT-5, OpenAI regains architectural simplicity while attempting to preserve user satisfaction.
However, this approach carries risk. If personality becomes a skin rather than a deeply embedded behavioral trait, users may continue to feel something is missing. GPT-4o wasn’t just friendlier by design choice; it behaved differently at a structural level. Subtle things — pacing, phrasing, empathy cues — mattered. Removing that option entirely may push power users and emotionally invested users toward third-party tools, open-source models, or API-based workarounds.
From a business perspective, retiring GPU-heavy legacy models makes sense. Maintaining multiple large models inside ChatGPT is expensive, inefficient, and harder to scale. From a trust perspective, though, repeated reversals and quiet sunsets risk eroding confidence. OpenAI says it will give notice — and technically, it has — but the lingering memory of last year’s surprise removal still hangs over this decision.
Ultimately, this move confirms OpenAI’s long-term strategy: fewer models, more abstraction, and heavier emphasis on customization layers rather than core diversity. Whether users accept that trade-off will define the next chapter of ChatGPT’s evolution.
🔍 Fact Checker Results
✅ OpenAI confirmed GPT-4o and three other models will be retired from ChatGPT on February 13.
✅ The models will remain available through the OpenAI API for developers.
❌ There is no evidence that GPT-4o will return again after this retirement.
📊 Prediction
ChatGPT will continue narrowing its model lineup, with GPT-5 becoming the dominant default while personality customization grows more prominent. Over time, user nostalgia for specific legacy models will fade publicly but persist privately, driving increased interest in API access, self-hosted alternatives, and open-source conversational AI.
🕵️📝✔️Let’s dive deep and fact‑check.
References:
Reported By: 9to5mac.com
Extra Source Hub (Possible Sources for article):
https://www.twitter.com
Wikipedia
OpenAi & Undercode AI
Image Source:
Unsplash
Undercode AI DI v2
Bing
🔐JOIN OUR CYBER WORLD [ CVE News • HackMonitor • UndercodeNews ]
📢 Follow UndercodeNews & Stay Tuned:
𝕏 formerly Twitter 🐦 | @ Threads | 🔗 Linkedin | 🦋BlueSky | 🐘Mastodon
