Issue: #124 Date: 2026-01-30T16:30:21-08:00

Summary

Yes, DuckDB supports graph queries through the DuckPGQ community extension, which implements the SQL/PGQ standard from SQL:2023. This allows you to perform graph pattern matching and path-finding operations directly within DuckDB without needing a separate graph database.

Key Findings

Native Graph Query Support

  • DuckDB doesn’t have built-in graph query support in core, but the DuckPGQ community extension adds SQL/Property Graph Query (SQL/PGQ) capabilities
  • SQL/PGQ is part of the official SQL:2023 standard, providing standardized syntax for graph operations

Core Capabilities

  • Pattern matching: Find specific patterns in connected data
  • Path-finding: Discover paths in your data, including variable-length paths
  • Property graph model: Label and assign properties to nodes and edges
  • Supports use cases like social networks, financial transaction analysis, and knowledge graphs

Technical Approach

  • DuckPGQ leverages DuckDB’s analytical capabilities to perform efficient, in-process graph analytics
  • Requires no external dependencies
  • Easy to install as a community extension
  • Allows you to work with graph data without exporting or managing a separate system

Current Status

  • DuckPGQ is actively maintained as a community extension (started as a research project)
  • Under ongoing development with recent activity as of October 2025
  • Some features may still be incomplete, so check documentation for your specific use case

Recommendations

If you need graph query capabilities in DuckDB:

  1. Install the DuckPGQ extension
  2. Review the SQL/PGQ documentation to understand the syntax
  3. Test with your specific graph patterns to ensure they meet your needs
  4. Monitor the extension’s GitHub repository for updates and known limitations

Sources