Error 400 when i try to upload a PDF file

[andreaslg0815]

Hi,

When I try to upload a PDF file, I receive the error message: "Error Processing File."

The logs show the following:

rag_api | 2025-09-05 07:00:38,917 - pypdf._reader - WARNING - Object 4 0 not defined.
rag_api | 2025-09-05 07:00:38,918 - root - ERROR - Error during file processing: Invalid object in /Pages
rag_api | Traceback: Traceback (most recent call last):
rag_api | File "/app/app/routes/document_routes.py", line 486, in embed_file
rag_api | data, known_type, file_ext = await load_file_content(
rag_api | File "/app/app/routes/document_routes.py", line 96, in load_file_content
rag_api | data = await run_in_executor(executor, loader.load)
LibreChat | 2025-09-05 07:00:38 error: Error uploading vectors The server responded with status 400: Request failed with status code 400
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_core/runnables/config.py", line 622, in run_in_executor
LibreChat | 2025-09-05 07:00:38 error: [/files] Error processing file: Request failed with status code 400
rag_api | return await asyncio.get_running_loop().run_in_executor(executor_or_config, wrapper)
rag_api | File "/usr/local/lib/python3.10/concurrent/futures/thread.py", line 58, in run
rag_api | result = self.fn(*self.args, **self.kwargs)
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_core/runnables/config.py", line 608, in wrapper
rag_api | return func(*args, **kwargs)
rag_api | File "/app/app/utils/document_loader.py", line 243, in load
rag_api | return loader.load()
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_core/document_loaders/base.py", line 32, in load
rag_api | return list(self.lazy_load())
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_community/document_loaders/pdf.py", line 305, in lazy_load
rag_api | yield from self.parser.lazy_parse(blob)
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_community/document_loaders/parsers/pdf.py", line 391, in lazy_parse
rag_api | "total_pages": len(pdf_reader.pages),
rag_api | File "/usr/local/lib/python3.10/site-packages/pypdf/_page.py", line 2164, in len
rag_api | return self.length_function()
rag_api | File "/usr/local/lib/python3.10/site-packages/pypdf/_doc_common.py", line 349, in get_num_pages
rag_api | self._flatten(self._readonly)
rag_api | File "/usr/local/lib/python3.10/site-packages/pypdf/_doc_common.py", line 1157, in _flatten
rag_api | raise PdfReadError("Invalid object in /Pages")
rag_api | pypdf.errors.PdfReadError: Invalid object in /Pages
rag_api |
rag_api | 2025-09-05 07:00:38,921 - root - INFO - Request POST http://rag_api:8000/embed - 400

Can anaybody help?

Thank you
Andreas

[dareima]

As far as I understand your error log, the document processing pipeline of LibreChat is using langchain, which relies on the pypdf library to read PDFs. The error is raised inside pypdf when it tries to process the /Pages structure of the PDF, and finds it broken or referencing a non-existent object.

I'd try something like the local PDF24 tools to process that PDF. You could try to compress it a bit for instance and then try re-uploading that new version to LibreChat again. It's also possible that PDF24 will already tell you there's a problem with that PDF when you try to use tools on it. Maybe that'll help as well.

[ihannsen]

LibreChat | 2025-09-05 07:00:38 error: Error uploading vectors The server responded with status 400: Request failed with status code 400

Seems you have something wrong with your vectordb, i think i can not contact it.
Check the connection from your librechat host to your vector db. Also double check your .env for vectordb parameters.
How did you setup your librechat? Docker?

[dareima]

LibreChat | 2025-09-05 07:00:38 error: Error uploading vectors The server responded with status 400: Request failed with status code 400

Seems you have something wrong with your vectordb, i think i can not contact it. Check the connection from your librechat host to your vector db. Also double check your .env for vectordb parameters. How did you setup your librechat? Docker?

Error 400 means bad request. It's not a connection issue. The PDF he is trying to process isn't living up to the expectations of a proper PDF file.

[andreaslg0815]

Thanks for your help.

I tried another PDF file and received the following log message:

LibreChat | 2025-09-08 08:21:41 error: Failed to load service key from file: /app/api/data/auth.json ENOENT: no such file or directory, open '/app/api/data/auth.json'
chat-meilisearch | 2025-09-08T08:21:41.897977Z INFO HTTP request{method=GET host="meilisearch:7700" route=/health query_parameters= user_agent=node status_code=200}: meilisearch: close time.busy=249µs time.idle=236µs
rag_api | 2025-09-08 07:56:34,171 - root - INFO - Unauthorized request with invalid token to: /embeddings, reason: Not enough segments
rag_api | 2025-09-08 07:56:34,172 - httpx - INFO - HTTP Request: POST http://0.0.0.0:8000/embeddings "HTTP/1.1 401 Unauthorized"
rag_api | 2025-09-08 07:56:34,178 - root - ERROR - Failed to store data in vector DB | File ID: b88c6552-06b6-4db5-87c1-906c70db40df | User ID: 67af0291b548c2cbb734ab2b | Error: Error code: 401 - {'detail': 'Invalid token: Not enough segments'} | Traceback: Traceback (most recent call last):
rag_api | File "/app/app/routes/document_routes.py", line 382, in store_data_in_vector_db
rag_api | ids = await vector_store.aadd_documents(
rag_api | File "/app/app/services/vector_store/async_pg_vector.py", line 69, in aadd_documents
rag_api | return await run_in_executor(
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_core/runnables/config.py", line 622, in run_in_executor
rag_api | return await asyncio.get_running_loop().run_in_executor(executor_or_config, wrapper)
rag_api | File "/usr/local/lib/python3.10/concurrent/futures/thread.py", line 58, in run
rag_api | result = self.fn(*self.args, **self.kwargs)
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_core/runnables/config.py", line 608, in wrapper
rag_api | return func(*args, **kwargs)
LibreChat | 2025-09-08 07:56:34 error: Error uploading vectors An error occurred while setting up the request: File embedding failed.
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_core/vectorstores/base.py", line 288, in add_documents
rag_api | return self.add_texts(texts, metadatas, **kwargs)
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_community/vectorstores/pgvector.py", line 564, in add_texts
rag_api | embeddings = self.embedding_function.embed_documents(list(texts))
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_openai/embeddings/base.py", line 590, in embed_documents
rag_api | return self._get_len_safe_embeddings(
rag_api | File "/usr/local/lib/python3.10/site-packages/langchain_openai/embeddings/base.py", line 478, in _get_len_safe_embeddings
rag_api | response = self.client.create(
rag_api | File "/usr/local/lib/python3.10/site-packages/openai/resources/embeddings.py", line 132, in create
rag_api | return self._post(
rag_api | File "/usr/local/lib/python3.10/site-packages/openai/_base_client.py", line 1259, in post
rag_api | return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
rag_api | File "/usr/local/lib/python3.10/site-packages/openai/_base_client.py", line 1047, in request
rag_api | raise self._make_status_error_from_response(err.response) from None
LibreChat | 2025-09-08 07:56:34 error: [/files] Error processing file: File embedding failed.
rag_api | openai.AuthenticationError: Error code: 401 - {'detail': 'Invalid token: Not enough segments'}
rag_api |
rag_api | 2025-09-08 07:56:34,179 - root - INFO - Request POST http://rag_api:8000/embed - 200

I should point out that the upload and file search had worked before, but the feature hasn't been used for quite some time. I can upload images and they are processed correctly.
The issue only occurs with PDFs.

[dareima]

Images are a different thing and are not processed by RAG.
This time your PDF went successfully beyond the pre-processing stage but it seems you have some issues in your RAG configuration.

These are the key problems:

Unauthorized request with invalid token to: /embeddings, reason: Not enough segments
Error code: 401 - {'detail': 'Invalid token: Not enough segments'}
openai.AuthenticationError: Error code: 401 - {'detail': 'Invalid token: Not enough segments'}

I suppose you are using OpenAI endpoints for AI models?
You might have a misconfiguration in your .env file regarding RAG and / or your OpenAI API key.

Can you post all respective lines from your .env file that start with RAG_ and possibly with EMBEDDINGS_ ?
Remember to redact personal information like API keys and such.

[andreaslg0815]

Yes, I'm using OpenAI.

RAG_OPENAI_BASEURL=http://0.0.0.0:8000
RAG_OPENAI_API_KEY=sk-proj-openai...
(#) RAG_USE_FULL_CONTEXT=
(#) EMBEDDINGS_PROVIDER=openai
(#) EMBEDDINGS_MODEL=text-embedding-3-small

[dareima]

The RAG_OPENAI_BASEURL doesn't look correct to me. This config looks like you'd want to re-route the official OpenAI endpoint to a local RAG system. Supposing you don't have that, give this simple configuration a try instead:

RAG_API_URL=http://host.docker.internal:8000
RAG_OPENAI_API_KEY=sk-your-openai-api-key-example

Don't use any other RAG environment variables except these two for your next test.
Here's the online reference from the LibreChat documentation -> RAG API Configuration

Remember to issue a docker compose down and then a docker compose up -d for the changes in the .env file to take effect.

[andreaslg0815]

That's it — I've removed the entry and it's working now. Thank you!