MySQL AI Enterprise Edition Released

Oracle has recently unveiled MySQL AI, a suite of new AI-powered features exclusively available in the MySQL Enterprise Edition, focusing on analytics and AI workloads for large-scale deployments. However, concerns have arisen across the MySQL community about the future of the popular open-source version, with fears of vendor lock-in exacerbated by recent internal layoffs. MySQL AI introduces vector storage and search capabilities, enabling businesses to build retrieval-augmented generation applications directly on MySQL without requiring a separate vector database. It also integrates with leading large language models, provides AI-driven query acceleration, and leverages in-database analytics to optimize workloads. According to Nipun Agarwal, Senior Vice President of MySQL Engineering at Oracle:

MySQL AI supports various native agent workflows, such as detecting financial fraud in stored banking transactions, monitoring in-store inventory, predicting product demand, and assisting customers with travel bookings. You can develop AI applications that access data from MySQL databases or file systems without data movement or complex integrations, and choose to migrate the same applications to MySQL HeatWave in the cloud.

The new AI engine comprises four core components: Generative AI, which allows users to extract accurate and context-aware information from documents in the local file system; a vector engine that enables developers to generate vectors from documents and store them in InnoDB’s vector storage; AutoML, which automates common training tasks like algorithm selection, data sampling, feature selection, and hyperparameter optimization; and finally, NL2SQL, which leverages large language models' text-to-SQL capabilities to allow developers to query database content using natural language. Since MySQL Enterprise offers native support for JavaScript stored procedures, developers can write JavaScript code using GenAI APIs that directly interacts with MySQL data. Additionally, MySQL Studio serves as a new unified interface for MySQL AI. Agarwal added:

MySQL Studio is the new visual interface for MySQL AI, offering an intuitive integrated environment that includes a SQL worksheet, a chat interface for querying documents in vector storage, and interactive notebooks for developing ML and GenAI applications.

The new interactive notebooks are compatible with Jupyter, allowing developers to import or share existing notebooks. This announcement follows recent layoffs and Oracle’s increasing AI focus on MySQL HeatWave, a managed MySQL Enterprise database service on Oracle Cloud Infrastructure (OCI), raising questions within the community about the future of MySQL as an open-source project. Patrik Backman, CEO of OpenOcean and co-founder of MariaDB, commented:

The original promise of MySQL was openness and freedom from lock-in (...) the features enterprises want most—analytics, ML, vectors—are increasingly locked into HeatWave. Staying with Oracle means deeper dependency on OCI.

There is growing concern within the community that Oracle is increasingly locking new features behind paid licenses while reducing support for the open-source version. Backman continued:

How much can Oracle reduce MySQL engineering before developer interest starts to decline? How long will large enterprises accept feature lock-in as the cost of continuity? (...) This isn’t just “Oracle being Oracle.” It’s a strategic shift. For enterprises, the cost of inaction may soon outweigh the cost of change.

Open-source database expert Mark Callaghan warned:

I hope Oracle continues investing in MySQL. It benefits the community and their HeatWave funnel, as the funnel stops working if users stop using open-source MySQL.

Trial downloads of MySQL AI are available on Oracle E-Delivery. The team has released a video introducing MySQL Studio and a MySQL AI demo showcasing travel booking using an MCP server.