I’m paying $20 a month for this service and you limit my messages WTF…reason to cancel. %40 were network errors or non responses. BS waste of money Google is free. And most answers are essentially are what Google returns as I usually have to ask the question 3x to get what I want.
I didn’t get close to 40 messages training my custom GPT and I got the limit alert. Something must be wrong, if so many users are reporting the same. I am looking into other options. Any recommendations to other no-code tools that I can build GPTs and integrate it with other apps?
Far too many messages here. But I agree… I was wondering if I was going to cancel my $20 subscription that I’ve had for so long now… Magic is going somewhere but not here
Same here. The cap is to small specially considering how much tweaking this thing needs to do what you want and remove mistakes and errors. What is the point of paying? I am seeing no advantage over the free one.
Same here, after few msgs get reach capped, you have 20 msgs… so its like 4 msg in 3 hourse, because you have to always fix his wrong answers gg
I have something very weird here. When I reached the limit, I waited until I could use 4.0 again. When I did, I was working on a php script, it suddenly came with this:
Verstehe, lassen Sie uns noch einmal betrachten, worum es in Ihrer Anfrage geht. Sie haben nach einem Bild gefragt, das eine Szene in einem antiken ägyptischen Tempel darstellt, in der eine Pharaonin, die eine Krone trägt und von einer Gruppe von Priestern und Priesterinnen begleitet wird, eine rituelle Handlung durchführt. Das Bild soll detailliert sein und Elemente wie Hieroglyphen, Säulen und vielleicht sogar einen Blick auf den Nil durch ein Fenster enthalten.
clear
468 / 5.000
Übersetzungsergebnisse
Übersetzung
I see, let’s review what your query is about. You asked for an image depicting a scene in an ancient Egyptian temple in which a female pharaoh wearing a crown and accompanied by a group of priests and priestesses performs a ritual act. The image is said to be detailed and include elements such as hieroglyphs, columns and perhaps even a view of the Nile through a window.
This had nothing to do with what we talked before. I asked, if it is not able to follow the chat before, but it refused to answer. I got the same:
I apologize for any misunderstanding and confusion that may have arisen. Let us address your concerns. How can I help you?
again and again.
We were not able to continue the work on my script until I again reached my limit. I am a bit upset. Is this some tactics to use up the the budget of prompts more easy?
The chat was in German language.
Same thing happening here. No way that was 40 messeges/3hours… skipping on the fact that the limit used to be 50/3hours. Now not only is it less but they don’t respect the limit anyway… I want to generate some content that I need for learning, and not only do I have to fix his answers all the time now, because his instruction following got absolutely stoopid (possibly with the context lenght increase i’ve noticed, possibly some bad context compression techniques).
So after trying to get answers that would be satisfactory I’ve got rate limited after only about I’d say 15-20 attempts at regenerating his answers with different prompting. Great job OpenAI, I’d love to see this website get scrapped by the rivals soon, maybe either the subscription will be less there, or they will actually value the customer.
@Ruphobia I agree with everything you said about training. I’ve heard the total cost sunk into ChatGPT4 is around one half billion $. So getting it for free, or almost free at $20/month, is a bargain!
BUT, I purchased this with the expectation that I could do useful work, because I demonstrated to myself that this was true before. Its an insane bargain, but I wouldn’t have bought in if the service was useless to me.
Now the service is useless to me. I don’t care what kind of bargain it is. I know only that they sold me a service that was useful to me, and now it is useless.
That is classic bait and switch.
This limitation is not useful at all. I understand that everyone has different requirements, but when I pay 20 USD / month i want to have a better tailored option. Instead of restricting it to 40 messages in 3 hours they could restrict it to 250 messages per 24 hours. That would made my day and I would be able to use the system much more efficient for myself. They could just make a button for the payed users to choose: which is fitting you better today. And we just choose it and it will be stored for 24 hours. after that it can be adjusted if required. This would give a much better way for users to be efficient and use ChatGPT in better ways.
It’s pretty low of OpenAI to not inform me about this restriction during the sign-up process. I’m shocked. This is pretty sketchy. I’m also confused of the talk about a supposed limit for 3.5 because I never even noticed one and I use it a lot… unless the times when it hangs unexpectedly without error was supposed to be some vague indicator that I reached some limit.
I suspect there will eventually be a class action against ChatGPT for this.
Oh no, I hope the $20/month isn’t a special “low availability” rate and if and when the restriction is lifted they also raise the price.
So when will this cap be lifted? I do not like having a limit on something that I am paying for. Why would you even have a cap for paid clients anyway?
Because the real “paid clients”, that are unlimited by only how much they want to spend, use the developer API, and their level of usage can rack up hundreds or thousands of dollars of AI data charges. Or contract million dollar annual dedicated instances.
ChatGPT is a personal consumer product that costs under $0.70 a day, while using AI inference that is computationally expensive and runs on $100,000+ servers equivalent to supercomputers a few years ago. And many that would abuse the access.
Here’s what a developer who wanted to make a chatbot that can make images would be paying to OpenAI:
Image models
Build DALL·E directly into your apps to generate and edit novel images and art. DALL·E 3 is the highest quality model and DALL·E 2 is optimized for lower cost.
Model Quality Resolution Price DALL·E 3 Standard 1024×1024 $0.04 / image Standard 1024×1792, 1792×1024 $0.08 / image DALL·E 3 HD 1024×1024 $0.08 / image HD 1024×1792, 1792×1024 $0.12 / image
$1 - which I can do in a handful of ChatGPT Plus requests.
On my account, I don’t even see an option for GPT4-Turbo, and certainly, getting cut off right in the middle of creating a GPT is pretty annoying. I suppose I could upgrade to “Teams” and I do intend to do that soon anyway as soon as my assistant is a little more familiar. But upgrading before then triples the monthly cost to $60 unless I shell out $600 upfront. I even purchased a third-party tool called “Team GPT” for collaborating. This cap really makes it hard to get anything done. I understand having a “fair use policy” but I truly believe it should be a lot more generous.
I just joined and it kicked me out while just setting up, I’m thinking of jumping out, this does not seem worth it
Very frustrating getting capped after 29 messages, all of which where spent dealing with sudden errors GPT-4 experienced accessing a .csv which it had been successfully accessing and reviewing with me earlier in the day. By the time I got it to access the file again so I could proceed, capped.
Not going to continue paying to see repeat errors if this is how caps will be handled.
I am using mobile internet, and unfortunately, it is very poor. I have to send messages several times, and as a result, I get blocked literally after 5-7 messages. Consequently, it is simply impossible to use your chat. It is unclear what I paid money for.
Apart from the limit itself, what’s annoying about the limit is how it works. There is no counter of how close you are, and when you hit the limit, it’s often for hours. There should be a leaky bucket model here, where you keep getting additional questions over time, and use from the bucket when you ask - and of course a counter that showed you how close you are to the limit.
Right now the “there was an error” counts as well, which is just downright silly.
Just upgraded to see the difference. play around to make my custom gpt in about 15 min - 20 min got kicked becuase of some limit. have to wait for 3 hours to most likely have a 15min - 20 min go again. RIP OFF! I feel i have been scammed out of my money. This is a terrible first impression. so much so i felt a need to make a account here to vent. Which is not a good sign for a paying customer. BAD IMPRESSION. Really dont care about the hype surrounding openai. Will definitely cancel my subscription once it runs out and not EVER renew. Will wait for the free open source ai or a MUCH MORE COST RESPECTFUL chinese or indian ai equivalent no doubt coming down the pipe 2 years from now. Really pissed and feel rip off right now. Served me right for giving good faith to a silicon valley company in the first place. PS - what little time i had with 4, had a sense of Google Gemini -condescending woke.