Comparing with all existing hashes in DB

Hi! It is works fine if I need to compare exactly two images. But what if I have 3 millions hashes in DB and need to compare my current hash with each other in database (find all similar from images that already exists). Is there any solution except direct comparing? May be it's possible to use something similar to distance function right in database?

No file, test, database engine, or entry point is named. Start by reviewing the package's hash-comparison API and clarify the target database, similarity threshold, and expected result before defining an implementable approach.

Written by the indexing model from the issue text.