Add API Endpoint for Posting Custom Parsed & Chunked Files to Knowledge Base

[SkypeForBusiness]

Description:
Currently, uploading knowledge into Open WebUI requires files to be processed internally using its built-in parsers and chunkers. However, this approach is not ideal in cases where users have complex file types or want to use their own custom parsers and splitters tailored to their specific use cases.

A quick use case:
A team processes complex scientific documents using a custom parser and chunker but still wants to make that knowledge available within Open WebUI to support RAG workflows. At the moment, they would either need to replicate their processing within WebUI or write directly to its database—both options are either too limiting or too invasive.

Proposed Solution:
Introduce a dedicated API endpoint that allows users to post a file along with pre-parsed and pre-chunked content. Open WebUI would then simply handle storing the data and integrating it into the knowledge base, without performing its own parsing or chunking.

Benefits:

• Enables use of custom parsers and chunkers outside the system

• Eliminates the need for direct database access

• Supports integration of complex or proprietary file formats

• Simplifies contributing external knowledge to the platform

• Makes Open WebUI more open and modular

Additional Context:
This should be relatively straightforward to implement: a REST endpoint that accepts a file and its associated chunks, validates them, and stores them appropriately. If the team is open to this feature, we would be happy to contribute and submit a PR for the implementation.