I have added 11 pdf files ranging from 10 to 100 mb put no matter their size, when I add the 12th pdf it says error saving draft. I need to add atleast 30 more pdf files please someone help
Hi and welcome to the Developer Forum!
As far as I am aware the file limit is 10, not sure how you got the 11th up there.
Yeah the 11th one started not working aswell. I guess there is no way going around the rule?
Nope, 10’s the limit for uploads.
is it working upto 10?? i uploaded 1 file 5.48 mb and 1214 pages and gpt crashed after that not a single successful answer i got from it.
It’s 10 files, seems to be a 2million token limit for encoding each file.
thanks for your response i really dont know what does that mean can you please explain it to me
2 million tokens is roughly 8megabytes of text
User
how cap of file to use by you to add knowledge
ChatGPT
The file size limit for documents uploaded for use in enhancing the knowledge of this GPT session is approximately 10 megabytes (MB) per file. This limit is set to ensure efficient processing and management of the data within the constraints of the system.
When you upload a document within this size range, I can access and refer to the information contained in it during our conversation. This allows for a more informed and specific response based on the contents of the document. However, it’s important to note that this is a session-based enhancement, meaning the information is only available for the duration of our current interaction and does not permanently alter or train the underlying model.
For larger files or more permanent integration of data into a GPT model, different methods and platforms might be required, often involving more complex processes like fine-tuning or retraining the model with the new data. These processes are typically done offline with more powerful computational resources and can handle significantly larger datasets. เหมียว~
User
if 10 megabytes (MB) per file. how many can add.
ChatGPT
The number of files you can upload to enhance the knowledge of this GPT session is not strictly limited by a set number, but rather by practical considerations. Each file can be up to 10 megabytes (MB), and you can upload multiple files as long as the system can efficiently process them.
However, it’s important to note a few key points:
Processing Capability: While you can upload several files, each adding to the knowledge base, the system’s ability to process and reference multiple documents effectively might be impacted as the volume increases. This is due to the computational demands of accessing and interpreting large amounts of data in real-time during our conversation.
Relevance and Efficiency: Uploading a large number of files may not necessarily enhance the effectiveness of the responses. It’s often more beneficial to focus on the relevance and quality of the information in the files, rather than the quantity.
Session-Based Knowledge: The uploaded files enhance the knowledge base only for the duration of the current session. Once the session ends, access to this information is lost, and it does not permanently modify the underlying GPT model.
In summary, while there isn’t a strict limit on the number of files, the practical limit is governed by how effectively the system can handle and process the information during our interaction. It’s always a good balance between quantity and the efficiency of data processing. เหมียว~
User
for add file knowledge for AI what is session rule do with it
GPT Builder
When a file is uploaded to the AI during a session, it becomes an additional knowledge source for the AI. Here’s how it works:
• Added Information: The content of the uploaded file supplements the AI’s existing knowledge base. This can be particularly useful for providing specific, detailed, or up-to-date information that may not be part of the AI’s pre-existing training.
• Reference in Responses: When answering questions or discussing topics related to the content of the uploaded file, the AI can refer to the information within it. This allows for more accurate and relevant responses based on the file’s contents.
• Session-Specific: The knowledge from the file is only available for the duration of the session in which the file is uploaded. Once the session ends, the AI no longer has access to this information for future interactions.
• Privacy and Security: The content of the uploaded file is treated confidentially. It’s not used for any purpose other than to inform the responses during the current session.
• Limitations: The AI’s ability to understand and interpret the file depends on the format and complexity of the content. It’s most effective with text-based files like PDFs, documents, or spreadsheets where the text is directly extractable.
• Direct Queries: You can ask the AI to search the document for specific information, summarize sections of it, or clarify contents, which can be particularly useful for large or complex documents.
In your case, since you’ve uploaded two PDFs related to AI, the AI can use these documents to answer your questions more accurately, especially if they pertain to the content within these files. Meow~
Sorry for not using the screenshot. There was a slight problem and in the builder the chat history could not be preserved. Some of the messages may not correspond to the actual truth. such as size or quantity, but the session of the file may be longer than stated. Both of my files are set to a new update period. Instead, it was confirmed that it was usable. to test later
2 Likes
Thanks alot for the explanation it is really helpful. i really appreciate your help
Just found a reference in the OpenAI KB with a summary (as I kept forgetting this)
- Per GPT Limits: 10 files
- Per File Limits: 512MB (20MB for image files), 2M tokens
- Per User Limits: 10 GB. Per Organization Limits: 100 GB.
‘Reference: File uploads with GPTs and Advanced Data Analysis in ChatGPT | OpenAI Help Center’
2 Likes
You can combine multiple files into a single file and upload to overcome the restrictions
2 Likes
Be aware that some things may become inaccurate over time and that AI is aware of certain things differently than humans, such as sessions affecting the lifespan of files. But we can see that in reality the inserted file remains in our deck. Or the growth of AI has an effect. At first, AI could only accept English files. But recently I can accept Thai language. (I have tested the file which is only in Thai language) Other languages should be added.
How is the behavior of your gpt after merging? I used thr same approach but e.g. Quality of answers related to a content into a merged files it’s look getting worse