Spring AI

Spring AI is an application framework for AI engineering from the Spring team. Its goal is to apply Spring ecosystem design principles — portability, modular design, and POJOs as building blocks — to the AI domain, addressing the core challenge of connecting enterprise data and APIs with AI models. It supports all major model providers (Anthropic, OpenAI, Microsoft, Amazon, Google, Ollama) across chat, embedding, text-to-image, audio transcription, text-to-speech, and moderation, with portable APIs for sync and streaming plus access to model-specific features.

Key APIs include ChatClient (a fluent API for chat models in the style of WebClient/RestClient), Advisors (reusable generative-AI patterns), structured outputs mapping model responses to POJOs, tools/function calling, observability, an ETL framework for document ingestion, model evaluation utilities, chat memory, and RAG support across a wide range of vector stores (Chroma, Milvus, Neo4j, pgvector, Pinecone, Qdrant, Redis, Weaviate, and more). Spring Boot starters for every model and vector store are selectable via start.spring.io. The project is Apache-2.0 licensed with over 9.5k GitHub stars.

1. Bootstrap a Spring Boot app at start.spring.io with Spring AI starters
2. Add your provider API key to application.properties
3. Build with the ChatClient fluent API