name: pgvector-python slug: pgvector-python url: https://github.com/pgvector/pgvector-python category: sdks-libraries brand: pgvector brand_logo: https://avatars.githubusercontent.com/u/101229709 images:
- https://opengraph.githubassets.com/1/pgvector/pgvector-python tags:
- sdk
- pgvector
- vector-store license: MIT featured: false
summary: | Python library for working with the pgvector PostgreSQL extension, enabling storage and similarity search of vector embeddings from Python-based AI/ML workflows and popular Python ORMs.
features: |
- Native Python client for pgvector, the PostgreSQL extension for vector similarity search
- Integrates PostgreSQL vector storage with Python AI/ML pipelines
- Support for popular Python frameworks and ORMs, including Django and SQLAlchemy (and likely other DB abstraction layers such as SQLModel)
- Provides tools to store, query, and compare vector embeddings directly in PostgreSQL
- Enables similarity search and nearest-neighbor queries on vector data
- Example code included (in
examples/) to demonstrate usage patterns and integrations - Test suite (in
tests/) for validating behavior and integrations - Distributed as a standard Python package (configured via
pyproject.toml) - Open-source under the MIT license
use_cases: |
- Persisting embedding vectors (from models such as sentence transformers or other embedding models) in PostgreSQL via pgvector from Python code
- Implementing semantic search or recommendation systems using vector similarity queries in Python applications
- Integrating vector search into Django- or SQLAlchemy-based web applications
- Prototyping and experimenting with pgvector functionality using the included Python examples
pricing: |
- Open source, MIT-licensed (no paid plans listed)
notes: |
- Full feature set and detailed integration examples are available in the repository’s README and examples directory.