- OpenAI says it will soon retire the GPT-4o model along with several others
- GPT-4o has a small but passionate user base that is unhappy with the move
- OpenAI feels that GPT-5.2 adds many features popular in GPT-4o
Back in August, OpenAI officially retired a number of its AI models, including GPT-4o. It was met with a swift backlash, forcing OpenAI to reinstate the model – but its time is finally up as GPT-4o is once again put out to pasture.
In an announcement on the company’s blog, OpenAI explained that “on February 13, 2026 … we will retire GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini from ChatGPT.”
The company added that a “special context” was needed regarding GPT-4o, noting that after GPT-5 was released and GPT-4o was initially retired, OpenAI received “clear feedback” from users. This revealed that many customers felt they “needed more time to transition important use cases, such as creative ideas, and preferred the GPT-4o’s conversational style and warmth.”
However, with the release of GPT-5.1 and GPT-5.2, OpenAI feels it has addressed most of these concerns. Furthermore, the company states that only 0.1% of users “still choose GPT-4o every day”, adding that it knows withdrawing from GPT-4o will feel “frustrating” and that “we did not make this decision lightly”. But doing so, it says, “allows us to focus on improving the models most people use today.”
The end of the road
Not surprisingly, given the reaction when GPT-4o was retired in August, the latest announcement has not gone down well with a section of ChatGPT users.
For example, user ClankerCore wrote on Reddit: “Time to go to change.org and start filling out petitions again. We brought back 4o last time. We’re bringing it back again.” For another paid ChatGPT subscriber, the decision was simple: “time to cancel.”
In a separate thread, another user described exactly why they wanted to keep using the GPT-4o: “Many users (myself included) feel that the GPT-4o offered something unique – not just in performance, but in personality, warmth and consistency. Some of us have built long-term creative projects, emotional support routines or studio workflows with this specific model. To lose it completely, or without reluctance, without even feeling left behind. Deeply disappointing.”
Still, this seems unlikely to change OpenAI’s mind. In the aforementioned blog post, the company said that “the vast majority of usage has shifted to GPT-5.2,” adding that this model has built-in features for “basic styles and tones like Friendly, and controls for things like Warmth and Enthusiasm.” It therefore appears that OpenAI is happy that GPT-5.2 meets the standards set by GPT-4o and that it will serve as a satisfactory replacement.
Whether that’s enough to appease the small but passionate user base of GPT-4o remains to be seen. However, with OpenAI apparently satisfied that GPT-5.2 addresses user concerns, it appears that this really is the end of the road for GPT-4o.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds. Be sure to click the Follow button!
And of course you can too follow TechRadar on TikTok for news, reviews, video unboxings, and get regular updates from us on WhatsApp also.
