This error happens very often: getNodeByIdOrMessageId - no node found by id: placeholder-request-…
Sometimes it just recovers itself and answers after 20-30 seconds.
Sometimes I need to induce it back to life.
But one time the chat died and I lost important information.
Windows App version is 1.2025.90.
I work on two main use cases.
Generating images for fun;
Generating code for an app project (python, sql, front and back end, vm and infrastructure).
Does anyone know if is there something that I can do to prevent this issue or if I am doing wrong? I am really concerned that this may impact my workflow.
Same is happening to me, retrying helps.
For me, the app helps reduce the load on chrome, but it’s painful to see this messages and see the app failing lot of times.
Same here. Has happened already 3 times this morning and several times yesterday. I use the mic input for text. It seems to be happening when the voice inputs a lot of information at once. i.e. smaller amounts of speech doesn’t seem to phase it as much. It is very annoying though. I have MS and memory issues. I can easily lose my thread (in my head) if I’m sitting there waiting for it turn it into text just for it to fail entirely. I’ve checked for updates / rebooted the app. I don’t use on browser but seeing the comments, it appears it is an issue there too.
I also get this it happens ALOT for like 1 month.
This is a MAJOR ISSUE why no one is fixing this? I’m also a subscriber and this is a product (defective) fault .
no idea why no one is fixing this i’m probably going to just use the browser version for now until we get an update on this. i didn’t have any of the same issues there.
I have the same problem regularly and it feels like something in network setup, but importantly - if I get this message and wait for some time, I receive full response once generation is complete. It only can’t display intermediate result
Yeah i’m having these issues too and it sucks because I do lots of coding and use it to troubleshoot and help bring my ideas to life with it’s feedback and give it some life, so I did some digging and it’s the problem with the User’s computer and the ChatGPT desktop app. Basically, the app relies on your computer’s RAM to handle conversations. When you’ve in really long or complex chats, it eats up more and more RAM because it’s not just forgetting the conversation mid way through. If your computer can’t keep up or if the app hits a snag in managing that memory, it throws that error. It’s like the app loses its place in the conversation because your computer is struggling to keep all that info loaded at once. This is why it’s super common with conversations that are really deep. This really sucks for members who are paying for it and are hitting these errors because it’s making it feel like they are not getting any proper explanation to fix it and are kinda being treated like free tier where it’s bare min. Some ways to prevent this error is to close it properly. Best way is through Task Manager or the backgrounds tab on the bottom right(Wins10, not sure abt 11) and waiting a bit for your computer to relieve from the load or saving your programs than doing a restart do wipe the memory.
And for the ones who question the website, it doesn’t happen on the web version as occasion as it does is because it runs on a server which has way more ram and proper hardware to handle much more strain and load than the average expensive or cheap PC in a household. It does use a bit of RAM on your computer’s end when going through the web version because it has to display all the information and because search engines use some RAM just to run (i.e the infamous chrome 0_o).
So yeah if this is happening, restart your PC after saving or do a reset on the ChatGPT app through proper closing and let your PC move free up that space. Yes it does make it slower to respond at first but that’s because it’s fetching the proper info from the conversation instead of having the important and older messages on it’s “mind”
Hope this helps and gives more clarity to the issue
In monkey terms(for ones who don’t like reading much:
ChatGPT app uses RAM for it’s train of thought. the train of thought isn’t forgetting older parts of your conversation. When you make long chat with it in one session it will hold onto all information. When you fully close it through Task Manager or background tasks, you’re resetting its train of thought. When you let RAM chill from such load and then start it back up, it has to think for a minute when responding to pick up relevant info and important messages from the conversation to respond properly. Problem doesn’t happen on the website rarely is because the website is being hosted on a big PC(server) with better hardware to maintain your conversation and others. When you use it on website it uses some ram to display everything in the conversation and for the search engine just to run. Big PC can hold many users at once with lots of conversations at once. So train of thought on your PC causes RAM to spike. Saving everything on PC then restarting it causes all RAM to be wiped and ready while just closing it properly will work but need time for PC to relieve from such load and to reset the train of thought. A good 5-15 mins is decent depending on your system capabilites
From other searches and issues showing up and now this- The ChatGPT App is horribly optimized and rather than providing an advantage for downloading and installing software, it gives a distinct list of disadvantages…
Also, like the 20 people above me asked… Why is there no actual reply to this?
Yeah I would say so, but then again the conversation “train of thought” memory is being ran through locally and not on OpenAI’s servers like the website version is. OpenAI most likely had the intention of having quick access to ChatGPT right there for sudden of thought ideas which would prove useful so the browser isn’t having to load and do all its checks to whatever browser you use then having to search and depending if you clear your history cookies or what not which would then log you out so they save the hassle for small things when you have a sudden thought you want to quickly ask about. They could also be wanting the main servers for their paying customers for their rewards of better queue and faster loading. While the benefits transfer also to the program it could be because they want to relieve and limit the stress from the servers for these benefits to actually take affect and having that wiggle room allows for it to have the speed it requires like an SSD. Having a good amount of space open at all times allows it to work the fastest it can because it has the room to just do it instead of having to not destroy data everywhere in the path to the computer. If your PC can handle it you wouldn’t be having this issue as often slim to none because you’re sharing resources between your pc and the servers. Also they could of tested to see the bare minimum of what the program could run on and just released it which is unstable because PC and laptop builds come in many different ways like a powerful boards but weak memory.