Solr VDB Provider¶
Apache Solr's dense vectors as a vector database for the AI module and AI Search. A site that already runs Solr for keyword search can keep its semantic search on the same service, with no second vendor.
docker run -d --name solr -p 8983:8983 \
solr:10.0.0 solr-precreate vectors
One core, any number of collections¶
The provider keeps every AI Search collection in one Solr core made from the
_default configset. It adds its vector fields and their types through the
Schema API, so there is no configset to copy or upload. It does not need
Search API Solr, and can use the same Solr server.
Filters before the search, scores AI Search expects¶
AI Search's conditions become Solr filter queries, which Solr applies before the kNN search, so a filtered search still returns the nearest matches. Scores are turned back into the similarity or distance that AI Search's thresholds expect. On Solr 10, scalar and binary quantization hold several times as many vectors in the same memory.
For AI agents¶
Solr on DigitalOcean is written for an AI coding agent working with a site's maintainer: it creates a Droplet, runs Solr behind HTTPS and basic authentication, and connects the provider to it. Agents can start from llms.txt.