Use Case
Your AI Doesn’t Need Another Database
MariaDB brings vector search, MCP and RAG into the database you already trust, so you can ground generative AI in live enterprise data without extra pipelines or copies.
What Is MariaDB for GenAI applications?
MariaDB builds vector search into its core database engine, so you can store embeddings next to relational data and ground LLMs in it with RAG. Your data stays in one secure platform, with no separate vector database or ETL pipeline to sync.
Capabilities
Key capabilities for GenAI applications
MariaDB covers the path from data to grounded answers: native vector search, RAG, Text-to-SQL, agent connectivity and framework integrations, all working with the relational data you already run.
Native vector search
Store, index and query embeddings alongside relational data in one engine. Run fast similarity searches and hybrid SQL queries without a separate vector database.
Hybrid search
Merge full-text keyword precision with vector similarity, then rank the combined results with Reciprocal Rank Fusion so users find the right answers faster.
MariaDB AI RAG
Deploy RAG on-premises without the plumbing. Native vector search and a REST API connect generative AI to your data, which never leaves your own MariaDB environment.
MariaDB Cloud AI retrieval
Fully managed RAG service for documents and databases. Parsing, hybrid retrieval and reranking handle your files, while Text-to-SQL answers from live data.
Framework integrations
Use MariaDB as the vector store for LangChain, LangChain.js, LlamaIndex and Spring AI, so developers can build with the tools they already know.
MariaDB supports a wide range of generative AI patterns out of the box
RAG assistants
Ground chatbots and support tools in your documents and knowledge bases. Retrieve accurate context and answer customer and employee questions with confidence.
Natural language analytics
Let business users ask questions in plain English. Text-to-SQL runs read-only queries on live operational data and returns immediate, accurate answers.
Semantic search and recommendations
Find similar items by meaning, not keywords. Combine vector similarity with structured filters like price and inventory in one SQL statement.
Get started
Start building with MariaDB
One real-time database platform you can trust, for everything you run today and everything you build next.
Frequently asked questions
No. MariaDB builds vector search into its core engine, so you manage embeddings and relational data in one place. That removes extra infrastructure, data synchronization and cost.
Retrieval-augmented generation grounds an LLM in your own data. Because models can hallucinate and lack business context, RAG retrieves relevant information first to help generate more accurate answers.
MariaDB runs AI workloads where your data already lives. Nothing moves between separate services, keeping your data within in your own MariaDB environment and reducing your attack surface.
Hybrid search combines full-text keyword matching with vector similarity. MariaDB merges both result sets with Reciprocal Rank Fusion, so users find exactly what they need faster.
Yes. The open source MariaDB MCP Server lets AI agents and assistants run read-only SQL queries and vector searches through the Model Context Protocol, using parameterized queries to improve security.
MariaDB Cloud AI Retrieval is a managed retrieval layer for RAG over documents and databases. Simple Retrieval APIs ground your AI apps in files and in live operational data.
They add MariaDB skills and an MCP server to coding agents such as Claude Code and Codex. Agents can read schemas, run SQL and test on a throwaway sandbox, with access limited to the connections and folders you allow.