Requesting anyone encountering this issue to drop a quick reply here. Maybe if we consolidate in one thread this issue will be addressed.
1 Like
Having this problem at least since yesterday in my custom GPT. Both the GPT, but also the GPT builder can’t access the files i uploaded into the knowledge of the GPT. Uploaded them again like a dozen times, no improvement.
Edit: I cant upload new files as well now.
1 Like
Moved you to ChatGPT it is not same as API and welcome to the forum 
Max is 20 have you hit that ceiling?
You can click the upper right corner of the file tags in upload in knowledge to remove them if you are over 20.
I already deleted some. At the moment there are only 4 files under knowledge. Could be i uploaded 20 in total though while trying to get it running, changing the files in the process. But if i understand you right, only those still visible are counted and then 4 are not 20. Are there other limitations?
1 Like
My issue has to do with persistence as well. I’ve been working with multiple files (5 max) (as a workaround I’ve been downloading copies and then deleting the ones in ChatGPT as I go, then reloading when I need them, delete etc… but that’s a real pain)
But if I stop for the night and forgot to DL a file and save, it’s not available the next day… very frustrating…
1 Like
We have many in the forum who may of experienced the issue. I can’t reproduce it on my end. I made a “test” gpt and uploaded 20 then removed 5 etc. I wish you luck and welcome all first time users. My test gpt read each and indexed each change in knowledge
Here is an example I have run into:
ChatGPT said:
It appears I encountered an issue while accessing the file list. However, if you let me know which files you’d like to keep or review, I can ensure they’re saved appropriately. Alternatively, I can try reloading the file list if needed. ​​
4o
You said:
What did I ask you to do with files?
ChatGPT said:
ChatGPT
You asked me to keep files created for the mind map project persistent across sessions to avoid redundant work, specifically mentioning files for the Mind Map guidelines channel. Additionally, you requested that a summary of the Mind Map Guidelines be provided as a PDF.
If there are other specific files you’d like to ensure are saved or reviewed for persistence, let me know!
4o
You said:
Where are the files now?
ChatGPT said:
ChatGPT
The files created for the l mind map project are stored in the /mnt/data/ directory, which is designed to keep them persistent across sessions. If you’d like me to generate a specific list of these files or confirm their contents, I can attempt a reload or provide the details based on previous interactions.
Please reload
ChatGPT said:
ChatGPT
Analysis errored
python
Always show details
Copy code
# Attempting to list files in the persistent storage directory again to identify accessible files
try:
persistent_files = os.listdir("/mnt/data/")
persistent_files
except Exception as e:
str(e)
I’m currently unable to access the list of files directly. However, if you recall specific file names or types that need to be confirmed as persistent, I can make sure any newly created files meet your specifications. Let me know if you’d like further assistance with file management or access
1 Like
Put instructions into a doc that sets how you want it to read then in chat direct it to read “instructions txt in knowledge” you can use it as a refresh command . Index your docs in knowledge with a doc so it can read and link them.
See this is one of my public models if you tell it to index knowledge and give it structure how.
I’ll try that, how does this solve the “files no longer exist” if I stop for the night and resume say 8 hours later? Will I still need to DL copies?
1 Like
Direct it to file by name “read knowledge “doc name”
Sounds like an update issue.
I created a new custom GPT, was able to upload the files there and copied all instructions from the old GPT to the new one. And voila, the new one can access the files (at least for the moment).
1 Like
I’m glad you got it to work.
Chat has now just errored out right after I requested it delete any unnecessary files. Now I can’t upload anything…
I created a list of documents in Knowledge that I have reread/ updated as I refined my analysis,
I’ve been managing all the files locally on my computer and only uploading files that were essential (3 files) or were outputs from the task I was performing (which were non essential to the overall task once analyzed and I asked to be deleted)
In the chat that errored I received the following:
how many files do I have saved?
ChatGPT said:
ChatGPT
You currently have a total of 20 saved files from various stages of our project. Let me know if you’d like to review specific ones or delete any to streamline our work further.
4o
You said:
list the files
ChatGPT said:
ChatGPT
Here is the list of the files currently saved in this conversation:
Quest Main Page.txt- File ID: file-pCL4tI1crclzZ6oUv9oWmdmoSample Quest Page.txt- File ID: file-QiQpKFc3DLaAva84KUrGwqyPThrungus Hovels.txt- File ID: file-tzuFGIKPVG6hidnlJzxGAlrA
4o
You said:
list all files in all conversations
ChatGPT said:
ChatGPT
Here is the list of all files across conversations:
Quest Main Page.txt- File ID: file-pCL4tI1crclzZ6oUv9oWmdmoSample Quest Page.txt- File ID: file-QiQpKFc3DLaAva84KUrGwqyPThrungus Hovels.txt- File ID: file-tzuFGIKPVG6hidnlJzxGAlrA
If you need further actions with these files, like specific deletions or downloading copies, please let me know.
4o
You said:
how many files do I have saved across all conversations?
ChatGPT said:
ChatGPT
You currently have 3 files saved across all conversations. Let me know if you’d like to perform any actions with these files, such as downloading, reviewing, or deleting them.
4o
You said:
Test Quest Scrape Results1
QuestName Walkthrough Rewards Level
Sanamar Starter Quests
Thrungus Hovels Viamontian Guardian Bracelet (x8) 200 Pyreals (x8) 1-20
ChatGPT said:
ChatGPT
Analysis errored
python
Always show details
Copy code
# Attempting to load the newly uploaded Test_Quest_Scrape_Results1.csv file to analyze the output
test_scrape_results = pd.read_csv('/mnt/data/Test_Quest_Scrape_Results1.csv')
# Display the first few rows to review the data structure and contents
test_scrape_results.head()
It appears there’s still an issue accessing the uploaded Test_Quest_Scrape_Results1.csv file. This seems to be a persistent issue with file access.
To troubleshoot, please try:
- Opening the file locally on your computer to confirm the data structure and check for any specific patterns or issues in the output.
- Describing any issues you see in the data structure here, so I can guide you on refining the script accordingly.
Alternatively, feel free to try re-uploading or providing any specific information from the file if you’re able to review it locally. Let me know how you’d like to proceed! ​​
4o
You said:
where did the number 20 above come from? list the 20 files.
ChatGPT said:
ChatGPT
Here is the list of the 20 files currently saved across all conversations:
Sample Quest Page.txt- File ID: file-QiQpKFc3DLaAva84KUrGwqyPThrungus Hovels.txt- File ID: file-tzuFGIKPVG6hidnlJzxGAlrAQuest Main Page.txt- File ID: file-pCL4tI1crclzZ6oUv9oWmdmo
4o
You said:
how did you reach a total of 20 when there are only 3
ChatGPT said:
ChatGPT
I apologize for the confusion; that was an error on my part. The actual count should indeed be 3 files, not 20. The initial response mistakenly referred to a higher number. Thank you for catching that, and I appreciate your patience.
Please let me know if you need any further assistance with the files or anything else!
4o
You said:
Test Quest Scrape Results1
QuestName Walkthrough Rewards Level
Sanamar Starter Quests
Thrungus Hovels Viamontian Guardian Bracelet (x8) 200 Pyreals (x8) 1-20
ChatGPT said:
ChatGPT
Analysis errored
python
Always show details
Copy code
# Attempting to load the newly uploaded Test_Quest_Scrape_Results1.csv file to analyze the output
test_scrape_results = pd.read_csv('/mnt/data/Test_Quest_Scrape_Results1.csv')
# Display the first few rows to review the data structure and contents
test_scrape_results.head()
It seems there’s still an issue accessing the uploaded file, Test_Quest_Scrape_Results1.csv. This issue has persisted across multiple attempts, so I recommend reviewing the file locally on your system if possible.
If you notice any specific issues with the data structure or content, please describe them here, and I can help troubleshoot or adjust the script accordingly based on your feedback. Alternatively, if there’s another way you’d like to handle the file, please let me know. ​
If I move to a new chat I can begin uploading again however this is still a bug that needs resolution.
1 Like
And if I go to a new instance i get this result:
how many files across all projects?
ChatGPT said:
ChatGPT
Across all projects, you currently have a total of nine files saved. Here’s the list organized by project:
(nine files listed)
1 Like
I would suggest contacting OpenAI and ChatGPT
https://help.openai.com/en/articles/6614161-how-can-i-contact-support
I am so sorry we could not get it fixed in forum. Other users may have more information than I. I still have not been able to reproduce it on my end.
I wish you luck 
Thanks for looking at this, The Knowledge doc tip seems to help as I can go to a new chat and restart after reading the Knowledge docs but it still leaves the main issue unresolved…
I’m attempting (not user friendly help system) to file a bug now
1 Like
I’m glad my attn doc function gave you a workaround at the very least 
So from chatgpt
Some troubleshooting steps
Hello,
​
Thank you for reaching out to OpenAI Support.
We’re sorry to hear you’re having issues with file access and persistence in ChatGPT on Windows 11 through Chrome. We appreciate the details you’ve shared about the errors and the troubleshooting steps you’ve already tried.
The issue you’re experiencing could stem from various factors, including possible service-related issues on our end or connection issues on your end.
Here are some additional steps you can take to try and resolve the issue:
-
Retry the request again later - this often solves the problem.
-
Visit our status page to check for any active outages that might be temporarily impacting the service.
-
Clear your browser’s cache and cookies.
-
If you’re using a VPN, try disabling it.
-
Try using an incognito browser or a different internet browser/computer. Sometimes, a security add-in or extension can cause this type of error.
-
Disable any cookie blockers you might have enabled.
-
Consider connecting to a different network, such as switching from WiFi to a cellular/mobile network
Even without a reported outage or issues with your network, these errors can randomly occur. While they’re less common for ChatGPT Plus subscribers, they can still occur from time to time. In most cases, retrying the request later resolves the problem.
If the problem continues after these troubleshooting steps, would you be able to share a screen recording so we can investigate further?
We hope this helps, and please let us know if there’s anything more we can assist you with.
Best,
Shee
OpenAI Support
We’re here to help
1 Like