vs

pgvector vs sqlite-vec

pgvector: The standard Postgres extension for vector similarity search with HNSW and IVFFlat indexes. sqlite-vec: Dependency-free SQLite extension for vector search that runs anywhere SQLite does, including WASM.

key differences

  • pgvector is built for larger datasets (≤ 10 TB) than sqlite-vec (≤ 100 GB).
  • Horizontal scaling is addon in pgvector and none in sqlite-vec.
  • Only pgvector has change data capture, stored procedures, materialized views, pub/sub, compression.
  • Only sqlite-vec has in-memory, runs in webassembly.
  • sqlite-vec can run embedded in your application process.

choose pgvector for

  • +Adding semantic search to an existing Postgres app
  • +RAG with permission filtering via SQL joins
  • +Recommendations next to product data
  • +Deduplication and similarity lookups

best at · Vector search, RAG, OLTP

choose sqlite-vec for

  • +On-device RAG in desktop and mobile apps
  • +Local-first semantic search
  • +Browser-based vector search via WASM
  • +Small server apps already on SQLite or libSQL

best at · Vector search, RAG, Embedded

overview
Primary model
Vector
Vector
Also supports
Relational
Relational
Type
Extension
Extension
Status
Active
Active
Developer
Andrew Kane and pgvector contributors
Alex Garcia (sponsored by Mozilla Builders)
First release
2021
2024
Latest version
0.8.7 (Oct 1, 2026)
0.1.9 (Mar 31, 2026)
License
Open source (permissive)PostgreSQL License
Open source (permissive)MIT or Apache-2.0 (dual)
Free tier
✓
✓
popularity
WhichDB Score
40.0
31.1
Rank
#44
#81
GitHub stars
23.5K
8.2K
Stack Overflow questions
91
–
scale & guarantees
Comfortable data size
Medium (≤ 10 TB)
Small (≤ 100 GB)
Horizontal scaling
Addon
None
Multi-region
Read replicas
None
ACID
Full
Full
Consistency
Strong
Strong
Schema
Strict
Strict
Architecture
Single node, Primary replica, Client server
Embedded
deployment
Self-hosted
✓
✓
Vendor-managed cloud
—
—
Serverless
—
—
Embedded
—
✓
Edge
—
✓
Clouds
AWS, Google Cloud, Azure, DigitalOcean, Heroku, Render
–
capabilities
Transactions
✓
✓
Joins
✓
✓
Secondary indexes
✓
✓
Full-text search
✓
✓
Vector search
✓
✓
Geospatial
—
—
JSON
✓
✓
Change data capture
✓
—
Triggers
✓
✓
Stored procedures
✓
—
Materialized views
✓
—
TTL / expiry
—
—
Pub/Sub
✓
—
Columnar storage
—
—
In-memory
—
✓
Durable persistence
✓
✓
Compression
✓
—
Encryption at rest
—
—
Role-based access
✓
—
Time travel
—
—
Branching
—
—
Online schema changes
✓
—
Streaming ingest
—
—
Offline sync
—
—
Runs in WebAssembly
—
✓
interfaces
Query languages
SQL
SQL
Client languages
Python, JavaScript, Go, Java, Ruby, Rust, C#, PHP
Python, JavaScript, Go, Rust, Ruby, C
Best for
Vector search, RAG, OLTP
Vector search, RAG, Embedded, Edge

pgvector vs sqlite-vec: FAQ

What is the difference between pgvector and sqlite-vec?

pgvector: The standard Postgres extension for vector similarity search with HNSW and IVFFlat indexes. sqlite-vec: Dependency-free SQLite extension for vector search that runs anywhere SQLite does, including WASM. pgvector is built for larger datasets (≤ 10 TB) than sqlite-vec (≤ 100 GB). Horizontal scaling is addon in pgvector and none in sqlite-vec. Only pgvector has change data capture, stored procedures, materialized views, pub/sub, compression.

When should I choose pgvector over sqlite-vec?

Choose pgvector for adding semantic search to an existing Postgres app, RAG with permission filtering via SQL joins and recommendations next to product data. Choose sqlite-vec for on-device RAG in desktop and mobile apps; local-first semantic search; and browser-based vector search via WASM.

Is pgvector or sqlite-vec more popular?

pgvector ranks #44 on the WhichDB Score (40.0/100) and sqlite-vec ranks #81 (31.1/100). Popularity is not fit: check the requirements that matter for your project.

Are pgvector and sqlite-vec open source?

Yes. pgvector is open source. License: PostgreSQL License. Yes. sqlite-vec is open source. License: MIT or Apache-2.0 (dual).