pgvector vs Turbopuffer: the honest comparison
The short answer
Edition v2026.Q3 · pricing and status verified 2026-08-06.
Choose pgvector for every startup that already runs Postgres and holds under ~10–50M vectors — which is roughly 95% of them. Choose Turbopuffer for multi-tenant AI products with millions of per-user namespaces — the exact shape of agent memory and per-customer document search.
On the elems table, pgvector holds seat #1 of the Vector Store element and Turbopuffer holds #2 — this is the closest call in the category, and the honest answer depends on which trade-off you can live with.
The case for each
1pgvectoropen source (Andrew Kane + community)
Free, PostgreSQL-licensed · runs in any Postgres — RDS, Supabase (free tier), Neon, Cloud SQLBest for Every startup that already runs Postgres and holds under ~10–50M vectors — which is roughly 95% of them.
Watch Vanilla HNSW strains past ~50M vectors — you'll want pgvectorscale (DiskANN) or a graduation plan. Index build memory and write-heavy churn need real Postgres operational skill; 'free' is not 'zero-ops'.
2Turbopufferturbopuffer, Inc.
Usage-based · Launch $16/mo min · Scale $256/mo min (HIPAA BAA, SSO) · Enterprise ≥$4,096/mo (single-tenant, BYOC)Best for Multi-tenant AI products with millions of per-user namespaces — the exact shape of agent memory and per-customer document search.
Watch Cold namespaces pay object-storage latency (hundreds of ms) on first hit. Closed source, no self-host below Enterprise BYOC, and a small team carrying very large customers. Revenue figure is single-source (Sacra).
pgvector vs Turbopuffer: side by side
Edition v2026.Q3 · verified 2026-08-06.
| pgvector | Turbopuffer | |
|---|---|---|
| Entry price | Free | $16/mo min |
| Open source | Yes (PostgreSQL lic.) | No |
| Deploy | In your Postgres, anywhere | Managed · BYOC at Enterprise |
| Hybrid (BM25+vector) | Via Postgres FTS | Yes (native BM25) |
| Sweet-spot scale | <10–50M vectors | 10M–100B+, many namespaces |
| Architecture | HNSW/IVFFlat in-database | Object storage + NVMe cache |
| Enterprise | Inherits your Postgres stack | SOC 2 · HIPAA · CMEK · 99.95% SLA |
What each side won't tell you
pgvector: Vanilla HNSW strains past ~50M vectors — you'll want pgvectorscale (DiskANN) or a graduation plan. Index build memory and write-heavy churn need real Postgres operational skill; 'free' is not 'zero-ops'.
Turbopuffer: Cold namespaces pay object-storage latency (hundreds of ms) on first hit. Closed source, no self-host below Enterprise BYOC, and a small team carrying very large customers. Revenue figure is single-source (Sacra).
If it's neither
The rest of the top five: Pinecone (managed classic) · Qdrant (open-source, filter-heavy) · Chroma (agent-native developer favorite). The complete field — 19 more tools including the graveyard — is on the element page.
This analysis is drawn from the Vector Store element dossier — the ranked top 5, the comparison matrix, and the complete field of 19 more tools live there, with every source. Data: vd.json (CC BY 4.0).
Every claim above is dated and sourced from the elems dossiers — 1,421 tools tracked across 58 categories, verified 2026-08-06, including the 276 we found dead, renamed, acquired, or sunsetting. Rankings are editorial, never paid — the charter.
Build your stack in 5 questions →