I am using gpt-4-1106-preview with retrieval enabled.
When using it with the playground, it can read the file just fine, but when i am uploading it through the API and then giving it the file ID as per documentation it says that it cannot access the file at all!
I asked the assistant to give me the file ID which it did and double checking it on the files section it does indeed exist, the file also says “Ready” next to the status, whatever that means, even so this problem persists.
The assistant swears that it doesn’t have the ability to access the file or its contents.
i have allocated multiple threads and it cannot even parse a 3 page pdf.
Meanwhile it is eating up my usage credits by writing massive paragraphs in apologies. Very disappointing to say the least considering that every chat is costing me a fortune here in India.
Please fix this OpenAI!