verifier.org

Vespa

the most sophisticated ranking engine here, with the steepest learning curve

$0.05 per vCPU-hour on the startup plan#products-where-retrieval#ranking-quality-is-the-d

a full search and ranking engine rather than a vector store — if ranking quality is your product, rank it second; if you just need rag, it's overkill.

vespa came out of yahoo and it shows in the best way: multi-phase ranking with real ranking expressions and machine-learned models is native, not an add-on. hybrid is not bm25 stapled to a vector index, it is a genuine tensor and ranking framework.

the pricing is the most transparent here — published per-vcpu, per-gb-memory, per-gb-disk and per-gb-gpu-memory hourly rates across four plan tiers, with unit prices declining as allocation grows and a discount for annual commitment.

the description above is ours, condensed from the ranking. pricing moves — check it on the vendor's own page before you rely on it.

pricing
$0.05 per vCPU-hour on the startup plan
website
vespa.ai
our verdict

a full search and ranking engine rather than a vector store — if ranking quality is your product, rank it second; if you just need rag, it's overkill.

we researched this category against vendors' own pricing pages and licence files. that is where this line comes from — not from the vendor, and not from anything they paid for.

more vector databases

Cloudflare Vectorize

vector index that runs on workers, beside cloudflare's own inference

#serverless#workers

Marqo

search engine that generates and stores its own embeddings for you

#embeddings-included

Upstash Vector

serverless vector database billed per request, with an http api

#serverless#http-api

pgvector

we checked thisfree — PostgreSQL License

the correct default for most teams, and it quietly got good

#almost-everyone-start-he#migrate-on-a-measured-th#not-a-vendor-s-table