Find answers from the community

Updated 8 months ago

Which vector database is the best choice

Which vector database is the best choice for LlamaIndex? I want strong similarity search
2
T
T
f
11 comments
You can check out the comparison here for feature support: https://docs.llamaindex.ai/en/stable/module_guides/storing/vector_stores/

I think the best one is Qdrant
what do you think about pg vector?
I think last time I checked the performance figures were very good
But haven't really used it myself
thanks for your answer.
what about document and index stores? in the documentation I see Mongo, Redis, Firestore. which one do you like? did you consider Postgress?

and the same question about the Index store.
I just want to choose one of them but need advice.
Any idea for Qdrant? It is said Qdrant doesn't support hybrid search here(https://superlinked.com/vector-db-comparison) but it does according to this(https://docs.llamaindex.ai/en/stable/module_guides/storing/vector_stores/)
Qdrant does seem to support both sparse & dense vector search, check https://docs.llamaindex.ai/en/stable/examples/vector_stores/qdrant_hybrid/
Any reasons you recommend Qdrant?
One of our teams is using Milvus, and we are being pushed towards deciding and sticking to one (my team uses Qdrant)
I'm pretty sure Qdrant has more features than Milvus.

Why I like Qdrant:

  • Great filtering
  • Supports everything I've needed
  • Has been reliable
  • Good console / UI
Honestly don't have a strong preference since I've only tried a handful. I was using Firestore but moving away from that now
Hey! I use Firestore for many of my applications just not for indexes. I am venturing into it. Can you please tell us more about your experience? Any specific reason to getting away from it?
Add a reply
Sign up and join the conversation on Discord