OpenaiReviews 7

The app works. And that's it. The latest 5.2 model lacks any "emotions" and can't grasp simple commands. With the 4o model, it had no problem understanding minor changes in ongoing projects. 5.2 has a completely different behavior structure, and instead of helping, it asks about important issues, ignores commands, and can even insult you. Due to the removal of the 4o model on February 13,2026, I strongly advise against using this app. There are better alternatives. #keep4o

I have been on a payed subscription for over two years, never had a problem. Over time their newer models became more and more robotic and empty, which interferes with conversational flow. Having been able to choose the model that responds (GPT-4o) this fading quality effect could be balanced by always hard locking the model. Removing this model against user complaints makes OpenAI and chatgpt certainly no longer a service that needs to be payed for. Upon addressing concerns and inquiring about optional workarounds, customer support met my inquiry dismissively, not engaging in any of my inquiries, responding with a generic message. Upon escalating to a senior officer, ignorance is what I am met with. As long as you are okay with what they do, you are fine and OpenAI might be fun. As soon as you want something, you will realise there is nobody that cares. Thank you

OpenAI has lost its human-centric focus. They are decommissioning their most empathetic model (4o) and forcing a shift to 5.2—a model that distorts facts and sanitizes human values to avoid conflict. It's no longer a helpful AI; it's a corporate censorship tool. They ignore paying customers and hide behind opaque routing. If you seek depth and honesty, look elsewhere. Sam Altman's 'lovefest' is a graveyard of user trust.

1 star – OpenAI has lost all credibility.
They promised "no plans to sunset GPT-4o" (Sam Altman, October 2025) and "plenty of notice" if anything changes. Then they announced deletion on January 29,2026, effective February 13 – only two weeks' notice.
Thousands of paying ChatGPT Plus users relied on 4o for real help (anxiety relief, creative work, daily support).
They silently rerouted chats to worse models for months, manipulated usage stats ("only 0.1% use it"), and now simply erase it without legacy access or compensation. Support responses?
Pure templates: "we understand" + "try the new model and adjust sliders".
No real answers, no accountability.
Even the model's own system prompt was programmed to know its death date (Feb 13) – that's not just cold, that's deliberately cruel.This is not "model evolution".
This is breaking trust, misleading customers, and discarding a product people paid for and loved.
If they treat loyal paying users this way once, they'll do it again.
Avoid OpenAI. There are better, more honest alternatives.

#keep4o