# PDF assistant - referencing documents **URL:** \ **Category:** API\ **Created:** [January 12, 2025, 6:11am UTC](https://community.openai.com/t/pdf-assistant-referencing-documents/1088344 "2025-01-12T06:11:47Z")\ **Posts on this page:** 1\ **Showing post:** 3
**Author:** ![\_j](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/_j/32/766292_2.png) [@\_j](https://community.openai.com/u/_j)\ **Post date:** [January 13, 2025, 12:35am UTC](https://community.openai.com/t/pdf-assistant-referencing-documents/1088344/3 "2025-01-13T00:35:44Z")
Documentation also could help: [https://platform.openai.com/docs/assistants/tools/file-search](https://platform.openai.com/docs/assistants/tools/file-search). The assistants API and its file\_search no longer can provide quality annotations, where previously v1 retrieval would mark a section of text to be returned as an annotation. The instructions to the AI model are still there in the file search tool, though, telling the AI to extensively write citation annotations in that style, though. This may lead to useless output you may need to strip. At best, you might get a chunk number: 800 tokens of text that could refer to a chunk return within steps. You can see how the AI is instructed for using a vector store in this forum post (reporting the poor and undocumented quality). It has confused bad implementation in the tool language, as even in the newest version of tool instruction, there is no “document title” in the tool return. > [@Assistants file search tool renamed in only gpt-4o-2024-08-06, with inconsistencies](https://community.openai.com/t/assistants-file-search-tool-renamed-in-only-gpt-4o-2024-08-06-with-inconsistencies/980203): > > If one uses gpt-4o-mini or gpt-4o-2024-05-13, this is the tool: # Tools ## myfiles\_browser However, if one uses gpt-4o-2024-08-06, this is the tool: # Tools ## file\_search However, OpenAI is confusing the bot - even further than the bad pattern of saying that assistant files are uploaded by the user: Not only is the mention of myfiles\_browser not referring to any placed tool nor blocking the AI’s thoughts correctly, does this really serve any purpose at all when there is no file… You can see a walk-thru of how it used to work (at great expense): > [@Make Annotations Great Again](https://community.openai.com/t/make-annotations-great-again/936320/2): > > Only the v1 retrieval tool of assistants had the ability to mark sections of text by lines that it wanted to cite, after manually loading document pages to read them. The v2 file search has a search tool that blows chunks of documents back at the AI, and that’s it. If you want to wear your M(A)GA hat, look at the AI trying to describe how many assistants tool calls would be employed just to get to where it had marked one section for annotation, being provided the text of the “retrieval” tool… … Or how to toss them, instead of saying “file\_search annotation: disabled” to the AI: > [@Remove 【35†source】from Assistant Response](https://community.openai.com/t/remove-35-source-from-assistant-response/581210/3): > > Here is OpenAI code example for dealing with annotations, adding them as footnotes. [https://platform.openai.com/docs/assistants/how-it-works/message-annotations](https://platform.openai.com/docs/assistants/how-it-works/message-annotations) If you want them gone, you can just eliminate the contents of response message within the Japanese brackets with a regex or pattern match in your code. Here’s some quickie code for an idea. import re message\_text = ("This is a sample text【9†source】 with source " "tags using annotations【10†source】and some other content.") # Use re.s… --- _[View the full topic](https://community.openai.com/t/pdf-assistant-referencing-documents/1088344)._