The deciding difference is operational: pgvector can run on infrastructure you control, Pinecone cannot. If that constraint is real for you, it settles the choice before anything else is considered.
Vector search inside the Postgres you already run.
Correct default under roughly ten million vectors. Reach for a dedicated store when it stops being.
Fully managed vector database. No index to operate.
| pgvector | Pinecone | |
|---|---|---|
| Licence | Open source | Proprietary |
| Pricing | open source | freemium |
| Self-hostable | Yes | No |
| Languages | Any language | Python, TypeScript |
| Install | — | — |
| Keys required | None | PINECONE_API_KEY |
The deciding difference is operational: pgvector can run on infrastructure you control, Pinecone cannot. If that constraint is real for you, it settles the choice before anything else is considered. pgvector: Vector search inside the Postgres you already run. Pinecone: Fully managed vector database. No index to operate.
pgvector can run on your own infrastructure. Pinecone is a hosted service only.
pgvector is open source (open source). Pinecone is proprietary (freemium).
Every page here answers to Accept: text/markdown and returns the same content at roughly a tenth the tokens. No separate site, no toggle — same URL.
curl -s -H "Accept: text/markdown" https://newagent.build/compare/pgvector-vs-pinecone