Cannot search pdf document using WebUI and Ollama

Users in the Ollama/Open WebUI community commonly report being unable to query PDF documents via the Open WebUI interface when using a locally hosted Ollama backend, with the model failing to recognize or access uploaded document content despite successful ingestion into the vector database. A key root cause identified is misconfigured embedding models — specifically, changing the embedding model without resetting the vector database renders existing vectors incompatible, breaking document retrieval. Additional contributing factors include OCR limitations for image-based PDFs, incorrect network/proxy configurations when Ollama and Open WebUI are hosted on separate servers, and the document context not being properly forwarded to the Ollama API at inference time.

Alternative to NotebookLM with no data limits
How do I know if an AI model could work locally on my computer?
Use the Same Model Across Ollama, LM Studio, Jan, and your Favorite Local AI Apps
Mac mini M4 48GB

Source: local-ai

Related