Drupal is a registered trademark of Dries Buytaert
Protected Pages 3.0.0 Major update available for module protected_pages (3.0.0). Commerce Core 3.3.8 Minor update available for module commerce (3.3.8). Search API HTML Element Filter 1.0.7 Minor update available for module search_api_html_element_filter (1.0.7). Layout Builder Reorder 2.0.1 Minor update available for module layout_builder_reorder (2.0.1). Ban 1.1.0 Minor update available for module ban (1.1.0). Field Formatter Range 2.0.0 Major update available for module field_formatter_range (2.0.0). Field Formatter Range 8.x-1.8 Minor update available for module field_formatter_range (8.x-1.8). Varbase Media Header 9.2.1 Minor update available for module varbase_media_header (9.2.1). Search API Solr 4.3.11 Module search_api_solr updated after 14 months of inactivity (4.3.11). Provus Mega Menu Module provus_mega_menu crossed 1,000 active installs.

This module allows Drupal to use MariaDB as a vector database for AI features, eliminating the need for separate external services. It leverages MariaDB's native vector data type and indexing to store and query AI embeddings directly within your existing database.

Native vector search for Drupal using MariaDB — no external services required.

This module enables MariaDB’s native VECTOR data type and indexing capabilities (available in MariaDB 11.7+) as a Vector Database (VDB) provider for the Drupal AI module. Since MariaDB is a drop-in replacement for MySQL and is widely used as Drupal’s primary database, vector embeddings can be stored and queried directly within the same database with full transactional support.

If your Drupal site already uses MariaDB, you can store your AI vectors in it by upgrading MariaDB to version 11.7 or later. If your site uses MySQL, you can replace it with MariaDB to enjoy the benefits of this module.

Features

  • Native MariaDB Vector Support: Utilizes MariaDB's built-in VECTOR data type and specialized VECTOR indexes
  • Multiple Distance Metrics: Support for Euclidean distance and Cosine similarity
  • Optimized Performance: Leverages MariaDB's modified HNSW (Hierarchical Navigable Small Worlds) algorithm
  • Seamless Integration: Works with the Drupal AI module and Search API
  • No External Dependencies: No need for separate vector database services
  • Operational Simplicity: No additional services, containers, or infrastructure layers

Post-Installation

  1. Create a Search API server using the AI Search backend
  2. Select "MariaDB vector DB" as the VDB provider
  3. Configure your embedding strategy and vector dimensions
  4. Create an index and add fields to the index
  5. Index your content

Your vector index is then ready for use with Drupal AI-powered features such as semantic search and RAG-based systems.

Additional Requirements

  • Drupal 10.2+ or Drupal 11+
  • MariaDB 11.7 or later (11.8 LTS recommended for GA support)
  • PHP 8.1+
  • PHP mysqli extension
  • Drupal AI module

Similar projects

Other AI Search VDB providers:

Commercial Support

This module is developed and maintained by Juan Martinez.

The module maintainer is available for professional consulting and architectural guidance related to production AI systems and enterprise Drupal implementations.

Activity

Total releases
3
First release
Feb 2026
Latest release
1 month ago
Releases (12 mo)
3 ▲ from 0
Maintenance
Active

Release Timeline

Releases

Version Type Release date
1.0.1 Stable Jun 17, 2026
1.0.0 Stable Mar 13, 2026
1.0.x-dev Dev Feb 22, 2026