I’m reaching out to escalate a persistent issue I’ve been facing with the ChatGPT platform. Despite being a paying subscriber on the GPT-4 Plus plan, I keep encountering the error message:
“Add file is unavailable”
whenever I attempt to upload files within the chat interface.
This is extremely frustrating, especially given that file upload is one of the core features promoted under the GPT-4 Plus subscription. The lack of access to this promised functionality undermines the value of the subscription and impacts my workflow significantly.
I believe it’s only fair that paying users receive full access to the services advertised. I would appreciate any assistance from moderators or engineers and would also like to know if others in the community are facing the same problem.
Please consider this an escalated concern, and I look forward to a prompt response or solution.
You’re not the only one. I’m a plus member too. I get “unknown error” and “Unable to extract text” errors even with the status page saying “resolved”. I’d suggest we wait until monday or tuesday until they have a full report of the causes so they can implement a proper fix. Right now they’ve only applied mitigations. I’ve tried every workaround or “fix” for these issues and I still get the same errors. .msearch and file parsing fails intermittently or silently fails in the background even after a “successful” upload. Again you’re not the only one. I’ve even tried different accounts and chatGPT hallucinates horrendously and receives many of the same errors.
Same here. I have tried different browsers. The only time ‘Add Files’ seems to be available is at the home page that too only when I open chatGPT for the first time in a day. When I move to continue with an already opened session, it becomes unavailable. Then even if I move back to the home page or open a new session, it remains unavailable.
Hi – I had exactly this problem and searched forums etc, - was very frustrating as a newly paid subscriber and wondered why I upgraded. However I just contacted support and got a quick reply that worked straight away and was so simple - I had to be in the chatgpt 4.0 model to upload - worked straight away! . Pity that ChatGPT could not diagnose itself! Mind you with all the false promises it has made me thus far about deliverables I am not surprised. From Support: 1. File Uploads (“Add files is unavailable” / button greyed out):
Make sure you’re using the GPT-4o or GPT-4 model in ChatGPT, as file uploads (including images and documents) are only available when you have one of these selected. If you are using a model such as “o1” or “o1-mini,” the upload feature is not supported—even on Plus.
On both desktop and mobile, you should see a “+” or paperclip icon next to the chat input when file upload is available. If this remains greyed out despite selecting GPT-4 or GPT-4o, you may be hitting a different technical issue..
I’ll also point out that that a lot of backend tools are non-functional. I only RECENTLY am able to upload documents LOCALLY (From my PC) Google docs fail regularly. however backend parsing, indexing, and .msearch and the like to extract text from documents are faulty and cause hallucinations. I have to tell chatgpt to use the python tool to manually extract text because the backend is unstable and faulty right now.
I am also experiencing the upload file error even while using 4o. (But yes, I have also noticed that the backend is really bad at document handling. They’re really trying to save on computing time… but it’s just wasting more of their resources)
It’s also causing more and more degradations on performance. the more they keep mitigating the tighter everything gets for people and another outage occurs like last time. I don’t understand why they just don’t FULL restore these functions by releasing an actual fix roll-back or introduce aouth systems for file and image uploads to alleviate files eating tokens and bandwidth every time they are uploaded.
Happening to me over and over again. Or something similar. My GTP just refuses to read files and hallucinates non stop no matter what file type I upload. It says it can read it then generates hallucinations 100% of the time
you can use the python tool for manual text extraction by telling chatGPT to do so. Though it eats up tokens once you start using it. Right now the backend tools have been silently failing and there have been throttles on a lot of users. you’re not the only one. For some of us we get fragmentation on just 10k-20k tokens when the context window SHOULD be 128k at least for paid accounts.
I have the same problem starting this morning in loading and deleting files in the projects knowledge base… no messagges, it only waits forever… completely stuck!!!
. Now resolved.