Comparing Agentic AI Frameworks for Java by Timo Salm / Sandra Ahlgrimm @ Spring I/O 2026

This talk at Spring I/O 2026 explores Agentic AI, emphasizing the transition from passive interactions to autonomous systems capable of reasoning, planning, and execution. The speaker compares various frameworks available on the JVM, including Spring AI, LangChain4j, and Embabel, highlighting their diverse approaches to implementing core agent capabilities such as reasoning loops, tool invocation, memory management, and orchestration. The discussion also addresses the role of the Model Context Protocol (MCP) in providing standardized access to data and tools across different agents. Attendees will gain practical insights into the trade-offs of these frameworks, aiding them in selecting the best solution for transforming JVM applications into fully autonomous agentic systems.