pgvector-lisp

Category: SDKs & Libraries
Brand: pgvector
License: MIT
Source: https://github.com/pgvector/pgvector-lisp

Lisp bindings and examples for using the pgvector PostgreSQL extension from Common Lisp, supporting multiple database libraries.

---

Features

Core Functionality

• Provides Common Lisp examples for integrating with the pgvector PostgreSQL extension.
• Enables using PostgreSQL as a vector store in Common Lisp applications.
• Supports vector operations via existing Common Lisp database libraries.

Supported Database Libraries

• Postmodern
  • Integration examples for using pgvector with the Postmodern PostgreSQL client.
  • Instructions for enabling the vector extension in PostgreSQL.
  • Example code for defining tables with vector columns and running queries.
• CL-DBI
  • Integration examples for using pgvector with CL-DBI.
  • Example code demonstrating how to work with vector columns via CL-DBI.

Example Workflows

• Embeddings with OpenAI
  • Example showing how to generate embeddings using OpenAI and store/query them in PostgreSQL via pgvector.
• Binary Embeddings with Cohere
  • Example demonstrating use of binary embeddings from Cohere with pgvector.
• Hybrid Search with Ollama
  • Example of hybrid search combining signals (e.g., using Reciprocal Rank Fusion) with pgvector.
• Sparse Search with Text Embeddings Inference
  • Example of sparse vector search using Text Embeddings Inference with pgvector.

Repository Contents

• postmodern.lisp – sample code for Postmodern-based integration.
• cl-dbi.lisp – sample code for CL-DBI-based integration.
• examples/ – multiple end-to-end examples for embeddings, binary embeddings, hybrid search, and sparse search.
• GitHub Actions workflows configuration under .github/workflows.

---

Getting Started

• Choose your database library:
  • Follow the Postmodern section for Postmodern-based projects.
  • Follow the CL-DBI section for CL-DBI-based projects.
• Use the examples in examples/ as templates for:
  • Setting up pgvector in PostgreSQL (enabling the vector extension).
  • Creating tables with vector (and possibly sparse) columns.
  • Inserting, updating, and querying vector data.

---

Pricing

• Not applicable. This is an open-source MIT-licensed library with no pricing plans.