Spring AI 1.0 is now generally available, and it is ready to help Java developers bring the power of AI into their Spring Boot applications. This release is the result of open collaboration and contributions across the Spring and Microsoft Azure engineering teams. Together, they have made it simple for Java developers to integrate LLMs, vector search, memory, and agentic workflows using the patterns they already know.

Why This Matters for Java Developers?

Spring AI 1.0, built and maintained by the Spring team at Broadcom with active contributions from Microsoft Azure, delivers an intuitive and powerful foundation for building intelligent apps. You can plug AI into existing Spring Boot apps with minimal friction, using starters and conventions familiar to every Spring developer. Whether you are building new intelligent features or exploring AI use cases, Spring AI and Azure have you covered.

Azure – The Complete AI Stack for Java Developers

Azure offers every essential component needed to build intelligent Java applications. At the core of this offering is Azure AI Foundry, which provides a unified platform for enterprise AI operations, model builders, and application development. With AI Foundry, Java developers do not need to train or fine-tune models. They can deploy foundation models themselves, interact with already deployed foundation models, or connect to models deployed by their AI engineers or data/ML engineers. Developers can connect to these deployed models, test prompt templates, inspect token usage and latency metrics, and embed model interactions directly into their Spring Boot applications. This platform combines production-grade infrastructure with user-friendly tools, helping developers operate AI-powered applications with confidence.

Here is how each piece fits into your Spring development workflow:

• Model-as-a-Service – Use Azure OpenAI for hosted large language models or choose from models available in Azure AI model inference, including offerings from Meta, Mistral AI, and DeepSeek. These models can be accessed directly using Spring AI starters, allowing your app to summarize, answer, generate, or assist through a simple, declarative API.


• Vector Databases – Embeddings and vector similarity are essential for semantic search and RAG. Azure provides multiple options: Cosmos DB with vector search support, Azure AI Search for advanced indexing and reranking, PostgreSQL with pgvector for SQL-based access, or Redis for in-memory speed. Spring AI integrates with these options to retrieve relevant context dynamically.


• Relational Databases – Use familiar relational databases like Azure SQL Database, PostgreSQL, and MySQL for structured data and transactional workloads. These remain the backbone for business logic, customer data, and system state that complement AI-generated content.


• Chat Memory – Track conversations across multiple turns using memory stores. Cosmos DB and Redis can persist chat memory, enabling your Spring-based agents to retain history, manage context, and respond in a more personalized way.


• RAG Workflows – Retrieval-Augmented Generation allows models to respond using external knowledge. Spring AI provides out-of-the-box support for RAG using vector stores and document loaders, enabling Java developers to build grounded, trustworthy interactions with minimal boilerplate.


• Orchestration with MCP – Build intelligent agents that call functions, retrieve data, and reason across steps. With Model Context Protocol (MCP) support, Spring AI apps can integrate seamlessly with external toolchains and invoke capabilities across multiple services. The MCP Java SDK is available through Spring Boot starters, simplifying the creation of both MCP Servers and Clients.


• App, AI Agent or MCP Server Deployment – Run your Spring AI applications, AI agents, and MCP servers on your choice of Azure compute: Azure App Service for managed deployments, AKS for containerized microservices, Azure Container Apps for serverless scale-out, or Virtual Machines for full control. Spring Boot simplifies packaging and deployment across these environments.

In addition to deploying applications built with Spring AI, you can deploy any MCP Server – regardless of the language stack – to Azure and interoperate with it using MCP Clients built with Spring AI. This ensures that Java developers can connect and collaborate with agents and services written by other teams using different technology stacks.

 Whether you are deploying an AI-powered web app, a backend intelligence service, or a full agentic workflow across services, Azure provides the flexibility, scalability, and operational reliability you need.

Fundamentals for Enterprise-Grade AI Applications

In production, AI features must align with enterprise requirements like security, reliability, and explainability. Here is how Spring AI and Azure together deliver on these needs:

• Security and Access Control – Ensure AI features respect role-based access policies. Use keyless or passwordless authentication for Spring Boot apps accessing Azure OpenAI, PostgreSQL + pgvector, Cosmos DB, and connections to Azure SQL, PostgreSQL, and MySQL – using managed identities and Entra ID. Integrate identity with Microsoft Entra ID and protect secrets using Azure Key Vault. Spring Security helps you enforce user-scoped interactions.


• Safety and Compliance – Use content filtering, prompt injection protection, and policy alignment to control model behavior. Azure OpenAI includes built-in safety filters. Combined with Spring AI’s prompt templates, you can enforce structured, policy-compliant interactions.


• Observability – Monitor model usage, token consumption, latency, and errors. Spring Boot Actuator and Micrometer provide metrics out-of-the-box. You can export these metrics to Azure Monitor and Application Insights for full-stack visibility.


• Structured Outputs – Use JSON or XML formats when integrating AI responses into downstream systems. Spring AI supports output parsing and schema validation, so generated content can drive actions within your application without post-processing.


• Reasoning and Explainability – Let applications show sources, highlight references, or explain decision flows. In domains like healthcare or finance, this transparency builds trust. Spring AI supports tool calling and multi-step workflows that help models reason and communicate clearly.

Tools, Demos and Learning Paths

• As always, you can start at start.spring.io, the easiest place to start any Spring Boot project, to quickly generate your app project using Spring AI, Web and Azure starters like OpenAI, AI Search, Cosmos DB, PostgreSQL + pgvector, relational databases, and Redis.

• End-to-End Demo – watch a hands-on demo of building enterprise AI agents with Java, Spring, and MCP.

• Get Started Today!

Resources: Azure QuickStart using Spring AI and Spring AI resources

Collaboration and Contributions

This GA release of Spring AI reflects the combined engineering and open-source commitment of Broadcom and Microsoft. Special thanks to Adib Saikali, Mark Pollack, Christian Tzolov, Dariusz Jędrzejczyk, Josh Long, Dan Vega, DaShaun Carter, Asir Selvasingh, Theo van Kraay, Mark Heckler, Matt Gotteiner, Govind Kamtamneni, Jorge Balderas, Brendan Mitchell, and Karl Erickson for their leadership, code, documentation, guidance, and community focus.

Get Started Today!

Everything you need is ready. Build your next AI-powered Java app with Spring AI and Azure.

Begin your journey here:  aka.ms/spring-ai