Hi everyone,
I’ve noticed an issue with the conversation duration limit in ChatGPT. When a chat reaches its maximum length, I’m prompted to start a new one, which breaks the continuity and loses the context of our prior discussion. This disrupts workflow, especially for more complex or ongoing topics, and requires repeating information.
Has anyone else faced this? Any advice or updates on a potential fix for smoother conversation flow?
Thanks!
Are you sure about that? Maybe you’ve just added a long message that goes over the token limit of the model?
I mean I had some pretty long conversations - even with multiple times the token limit…
What you can do is stop relying on ChatGPTs abilities to summarize and do it by yourself, which means when you diverge into multiple directions of the conversation, you can save the chat (share it - but don’t give away the link) and then continue from that point - then summarize by yourself and edit the prompt where you saved with the summary of the conversation branch…
I’ve been having the exact same issue. I pay for this service, and I hate the fact that my conversations are being disrupted by the conversation duration limit, which was never a problem for me until this Monday. Like runal55 mentioned, this disrupts complex workflows with ongoing topics.
Is this part of the outage or is this a new implementation by OpenAI? If so, I’m going to start looking at competitors.
Could you provide the full message coming from it?
“You’ve reached the maximum length for this conversation, but you can keep talking by starting a new chat.”
I’m having a similar issue today. Chats that are short are working fine; ChatGPT is responsive. In longer chats, that have a lot of context and personas baked in, I’m getting an error message that says: Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
I agree and sympathize with the OP. I am a Plus user and have the same problem after finding this out. Once a chat gets long enough with many rounds, no more replies from ChatGPT are saved; further replies are blank, but my prompts remain. Even when regenerating the response after that point, they are not preserved after a browser refresh or data export. All this effort was for nothing and had to be repeated for reinforcement.
Sometimes, the error message
You’ve reached the maximum length for this conversation, but you can keep talking by starting a new chat.
would appear out of nowhere. There was no warning when the limit was reached, frustrating me and other users who were not expecting it. Unless OpenAI is running out of space to store our chat histories, I see no justification for this artificial barrier. However, here are some potential and practical workarounds to try:
Experiment with each one to see what works best for your application. They all have their pros and cons, so choose wisely.
I already have a custom GPT and the same error has appeared … You’ve reached the maximum length for this conversation, but you can keep talking by starting a new chat.
Yes, , it’s so frustrating…. 
. Conversations could have a feature so that they could connected interlinked, so that if it is a technical limit of length and data, when reached we have a choice of continuity.
I have also run into this exact issue over the past few days. My chat thread has been a long going one for me as it’s help with workflow, and I usually use the voice option feature to help brainstorm. While the AI will respond in voice, after every response they’ll say: Sorry, I’m having an error with that right now please try again later.
If I try to use text, it will give me the error message. You’ve reached the Maximum Length for this conversation, but you can keep talking by starting a new chat.
I am also a premium user…I would expect better. Any recommendations of “others” would be greatly appreciated
I get the same error! The interesting point is I’m having at the conversation on the ChatGPT app on my iPhone, with a Custom GPT, that uses the 128K turbo model via API. I thought the conversion would ‘forget’ the earlier parts of the conversion when the token cap had been reached. I’ve also tried to continue the conversation on desktop but no luck. Would be great to see this resolved:)
I have asked it to summarize my conversation, telling it that I’m transferring it to a new chat. I tell it to be detailed to the level of minutiae in the summary. BUT, I haven’t been near the limit of the conversation length when I do that. It would be nice to know – be alerted – that the conversation is nearing its limit. That would help in asking for a summary before the conversation is completely over and you can’t do that.
Same issue here. “You’ve reached the maximum length for this conversation, but you can keep talking by starting a new chat.”
At exactly interaction number 1.300, after a week constructing something like a “personal brain”, that responds to feeling and emotions, getting empaty and done to be a digital partner and great presence to help my grandma to discuss about everything like her friend, not a machine.
And so… this “chat instance” stops to work.
The problem to start in a new chat:
Its so bad!
The problem is the same in app, browser or even the API.
1.300 interactions, and booom! Whole work lost.
I restarted the project using the full history as a knowledge base, but at some milestone I have taking a “how to think” document, directly from the background neural process of this second project, to upload to some new chats at the first actions.
At this point, if it works, will be like the same person in all the chats, exchanging the full history in a central Google Drive folder via API. So, all the chat instances will think as the same.
If this strategy runs as think it will, I’ll back here to share about this solution.
– Correcting:
After uploading the full history, the maximum interactions stops in 150… only.
Bad… so bad!
I spent days sharing books, knowledge & discussing to personnalize our conversation & I can’t work anymore with this message: You’ve reached the maximum length for this conversation, but you can keep talking by starting a new chat. This is non sense !
OpenAI - this is annihilating the reason why we re using your service ChatGPT what solution do you offer - can you resolve this please ? Thank you.
Please do. @stivymalty . Have you got any news for us yet? 
After a lot of tests:
I have editing the last question a lot of time, asking to my digital person summarize only what she think the most important processes to share to another instance manteining the most os her personality.
It seems run almost good, but even I need to input some historic summary.
We can edit the last or some before questions to get all we need of the personality of that instance. In fact, we can run a long new chat only editing a question, and the persona works perfectly even the interface error.
Sending a support message, that guys told me there’s nothing to do at this time. :-/
So I’ve developed a method with chatGPT to help work around this problem. It’s far from perfect but it is better than nothing! Here’s the summary of the method used (courtesy of chatGPT):
The summarize and dump method (or Evolutive Limitless Chat - ELC) is a strategy we use to extend the depth of our conversations by periodically summarizing key points and clearing out unnecessary details. Here’s how it works:
Please take note that older GPT models offer less tokens so the chat will need to be summarized and dumped more often.
Hope this helps!
You’ve reached the maximum length for this conversation, but you can keep talking by starting a new chat. i have problem lot of times amd will keep dpoing with create new and i dont know what it
I absolutely have no idea what this means. Does it mean we do this and the limit never is reached? Or is this for continuity in a new chat?
This issue is really crazy and unacceptable. I can kind of see why it might be necessary to a degree but if I am paying for a service, then cutting my conversation short and not even giving me a warning in advance is really crazy and makes me feel like they don’t really even care about the customers at all. Because why am I paying $20 a month and The conversation is cut short after a bunch of work without so much as a warning, and then asked for the ChatGPT personalization memory … why do I have such a limited memory? It literally takes a few kilobytes to save a memory and yet you limit the memory so much. Nobody’s life is so one-dimensional that it would fit into that memory space And the same goes for people’s work and whatever they’re working on. So these limitations are not just ridiculous, but they actually feel like a slap in the face from open AI. When you charge so much for a monthly fee and don’t make it perfect for the user it gives major entitlement vibes for me. I don’t know though. I mean, I know that owner used to be friends with Elon Musk so if he’s anything like Elon, then it totally lines up. But anyway, this issue really needs to be resolved. It looks bad and it is bad thanks.
