Postgres extensions - verified
pgvector (Postgres extension)
pgvector is listed under postgres extensions (deployment type: database extension). Open source; no price. Hosting cost depends on the Postgres provider. Max dimensions: 16,000 (stored vector type; HNSW and IVFFlat index up to 2,000 dimensions for vector, 4,000 for halfvec). Stated: hybrid and sparse search. Read from the vendor's own pages on October 18, 2026.
Read from github.com, raw.githubusercontent.com. Docs - License - Compare every vector database
Status
- Availability label
- Not stated
No release-status label is stated in the README.
How pgvector charges
Every price pgvector states, in its own unit, with the quote and where it comes from. Nothing is converted.
Open source; no price. Hosting cost depends on the Postgres provider.
Free tier and limits
- Free tier
- Not stated
Open-source extension; no hosted free tier applies.
- Limits
- Not stated
No row-count limit is stated in the README.
Search and indexing
- Max dimensions
16,000
Applies to: stored vector type; HNSW and IVFFlat index up to 2,000 dimensions for vector, 4,000 for halfvec
“Vectors can have up to 16,000 dimensions.”
raw.githubusercontent.com “- vector - up to 2,000 dimensions - halfvec - up to 4,000 dimensions - bit - up to 64,000 dimensions”
raw.githubusercontent.com - Distance metrics
L2, inner product, cosine, L1, Hamming, Jaccard
“L2 distance, inner product, cosine distance, L1 distance, Hamming distance, and Jaccard distance”
raw.githubusercontent.com - Index types
HNSW, IVFFlat
“An HNSW index creates a multilayer graph.”
raw.githubusercontent.com “An IVFFlat index divides vectors into lists, and then searches a subset of those lists that are closest to the query vector.”
raw.githubusercontent.com - Hybrid and sparse search
Yes, with Postgres full-text search
“Use together with Postgres full-text search for hybrid search.”
raw.githubusercontent.com - Metadata filtering
Yes, via ordinary SQL WHERE clauses and JOINs
“Plus ACID compliance, point-in-time recovery, JOINs, and all of the other great features of Postgres”
raw.githubusercontent.com - Multi-tenancy
- Not stated
Not described in the README; standard Postgres schemas and row-level security apply.
Built-in embedding and reranking
- Built-in embedding
- Not stated
The extension stores and searches vectors; embedding generation is not part of it per the README.
- Built-in reranking
- Not stated
Not part of the extension per the README.
Deployment, license and SDKs
- Regions and clouds
- Not stated
Runs wherever Postgres runs; regions do not apply.
- BYOC and self-host
Open-source Postgres extension
“Open-source vector similarity search for Postgres”
raw.githubusercontent.com - License
PostgreSQL License
“Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group”
raw.githubusercontent.com - SDKs
Any language with a Postgres client
“any language with a Postgres client”
raw.githubusercontent.com
Vendor-stated claims
Performance, scale and SLA figures as pgvector states them. Not measured.
No performance, scale or SLA figure is recorded from the pages checked.
Research notes
Supports single-precision, half-precision, binary and sparse vectors per the README.
FAQ
How much does pgvector cost?
Open source; no price. Hosting cost depends on the Postgres provider.
What is the maximum vector dimension in pgvector?
16,000 (stored vector type; HNSW and IVFFlat index up to 2,000 dimensions for vector, 4,000 for halfvec), per the vendor: "Vectors can have up to 16,000 dimensions."
Does pgvector support hybrid search?
Yes, with Postgres full-text search, per the vendor: "Use together with Postgres full-text search for hybrid search."
Does pgvector have built-in embedding or reranking?
Embedding: not stated. The extension stores and searches vectors; embedding generation is not part of it per the README. Reranking: not stated. Not part of the extension per the README.
Sources
Last verified from each vendor's own docs, pricing pages, READMEs and license files; every quote was string-matched against the page that day. This is a dated snapshot, not a benchmark: prices are shown in each product's own unit and change without notice, so check the linked source before you commit.
Newsletter
Know when a vector database changes its pricing
One email per week. New model releases, retrieval tooling and pricing changes and what they mean for what you build.