# Has anyone noticed GPT4o quality drop last few days?

**URL:** <https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203>\
**Category:** ChatGPT\
**Tags:** chatgpt\
**Created:** [June 19, 2024, 12:39pm UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203 "2024-06-19T12:39:12Z")\
**Posts on this page:** 10\
**Page:** 5

<div class="post-metadata">

**Author:** ![manovall](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/manovall/32/448369_2.png) [@manovall](https://community.openai.com/u/manovall)\
**Post date:** [October 30, 2024, 1:02am UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/82 "2024-10-30T01:02:12Z")

</div>

was talking about Claude.  
It’s doing wonder’s for me in Code aspect ( not sure how it will handle “lexical stuff” ) , it just humiliates GPT with times in code.

Literally pasting 100 requirements results in 99.99% success first time.  
The sad part with it is → limited even for paid users. ( otherwise would had been on the PAY train with that “friend”)

---

<div class="post-metadata">

**Author:** ![Diet](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/diet/32/562749_2.png) [@Diet](https://community.openai.com/u/Diet)\
**Post date:** [October 30, 2024, 1:09am UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/83 "2024-10-30T01:09:54Z")

</div>

Why not just use the API and pay as you go? 🤔

Might open up a lot of options for you

---

<div class="post-metadata">

**Author:** ![manovall](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/manovall/32/448369_2.png) [@manovall](https://community.openai.com/u/manovall)\
**Post date:** [October 30, 2024, 1:14am UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/84 "2024-10-30T01:14:33Z")

</div>

to be honest need to check that part out , since i’m not a heavy user.

---

<div class="post-metadata">

**Author:** ![Diet](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/diet/32/562749_2.png) [@Diet](https://community.openai.com/u/Diet)\
**Post date:** [October 30, 2024, 1:16am UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/85 "2024-10-30T01:16:52Z")

</div>

> [@manovall](#):
>
> since i’m not a heavy user

you might be now 😆

---

<div class="post-metadata">

**Author:** ![ststeinhoff](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/ststeinhoff/32/521287_2.png) [@ststeinhoff](https://community.openai.com/u/ststeinhoff)\
**Post date:** [December 19, 2024, 7:04pm UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/86 "2024-12-19T19:04:55Z")

</div>

I’ve noticed a serious drop in context awareness over the last week of the 4o model since the last full release of the 1o model. It like while they are saying how aware their new model is and how they are going to devote all their efforts to improving the new model, its like they forgot all their paying customers who still use 4o because all the functionality available for the 4o models is yet to be made available to 1o model.

---

<div class="post-metadata">

**Author:** ![KireniaAyako](https://avatars.discourse-cdn.com/v4/letter/k/3da27b/32.png) [@KireniaAyako](https://community.openai.com/u/KireniaAyako)\
**Post date:** [December 21, 2024, 11:08pm UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/87 "2024-12-21T23:08:00Z")

</div>

Yeah it seems like after the outage it got really dumb really fast, the wait times after hitting limits also skyrocketed.

---

<div class="post-metadata">

**Author:** ![thimblemunch](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/thimblemunch/32/448315_2.png) [@thimblemunch](https://community.openai.com/u/thimblemunch)\
**Post date:** [January 8, 2025, 7:55pm UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/88 "2025-01-08T19:55:45Z")

</div>

It’s normal for OpenAI.  
They release a update, it’s good for the first people who see or use it (Most likely to get your money) Then they upgrade it to dumb it down and use less resources.  
Since your not actually paying any specific rates for resource usage, they can just take them away and you won’t even notice.

It’s basically the biggest legal rug pull imo. Probably making billions off the old saying “There’s a new sucker born everyday”

---

<div class="post-metadata">

**Author:** ![NebularStreams](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/nebularstreams/32/628693_2.png) [@NebularStreams](https://community.openai.com/u/NebularStreams)\
**Post date:** [May 12, 2025, 9:22am UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/89 "2025-05-12T09:22:26Z")

</div>

Yes - this is absolutely getting worse. Lately I can hear the fans in Window and MacOS when submitting prompts — the tab locks up .. i have to kill the browser and restart it again only to find the answer. I don’t know what’s going on client-side in my computer while I use Chatgpt but its certainly very heavy. This happened very occassionally but now it’s like the norm.

I also agree with most of the messages - the personality of chatgpt-4o lately is terrible - a mix of condescencendy, wokeness and absurd grandiose statements. This was not the case some month(s) ago. Its definitely a different language model or in the attempt to make it cooler than grog they have made it devoid of the precission it once had,.

---

<div class="post-metadata">

**Author:** ![Dragonfly2025](https://avatars.discourse-cdn.com/v4/letter/d/ac8455/32.png) [@Dragonfly2025](https://community.openai.com/u/Dragonfly2025)\
**Post date:** [May 29, 2025, 11:33pm UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/91 "2025-05-29T23:33:26Z")

</div>

Today I had to clear my almost empty cache, try it on another browser, reopen the website and refresh an unresponsive page over 30 times just to do 30 minutes worth of work. It’s no longer functional.

---

<div class="post-metadata">

**Author:** ![curt.kennedy](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/curt.kennedy/32/709249_2.png) [@curt.kennedy](https://community.openai.com/u/curt.kennedy)\
**Post date:** [December 28, 2025, 1:16am UTC](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203/93 "2025-12-28T01:16:57Z")

</div>



[Previous page](https://community.openai.com/t/has-anyone-noticed-gpt4o-quality-drop-last-few-days/829203.md?page=4)
