from all the hosted vector stores llama

to start with, you could just use our in-memory simplevectorindex, doesn't use a hosted vector provider

I need hosted because I am using a serverless api and don’t want to spin up a server always running would be more expensive to start with than a hosted vector

I see. Hmm it's a good question, off the top of my head the main "hosted" versions we have are pinecone and weaviate atm

the only one with a good enough free plan is pinecone so I have to start with that but I’ll use a local one to test custom stuff also, maybe faiss