Spring AI 2.0 Goes GA: Unified Tool Calling, MCP Changes, and the Latest from OpenAI/Anthropic

Spring AI 2.0 has shipped as a general availability release, and the latest episode of the Spring Office Hours podcast (S5E18) breaks down what changed. Dan Vega and DaShaun Carter cover the new unified tool calling model, progressive tool discovery, and what's different in MCP (Model Context Protocol) support. The release is built on Spring Boot 4 and Spring Framework 7, so if you're still on older versions, plan your upgrade path.

What's New in Spring AI 2.0

The headline feature is a unified tool calling model. Instead of juggling multiple abstractions for different AI providers, Spring AI 2.0 provides a single interface for tool invocation across OpenAI, Anthropic, and others. Progressive tool discovery means the framework can dynamically add tools as the conversation evolves, rather than requiring a static list upfront. This is particularly useful for complex agent workflows where the set of available functions depends on previous responses.

MCP support has been overhauled. The protocol that lets AI models interact with external tools and data sources now has tighter integration with Spring's programming model. Expect less boilerplate and more idiomatic Spring configuration for defining MCP servers and clients.

OpenAI and Anthropic Updates

The podcast also touches on recent model releases. OpenAI previewed GPT-5.6 Sol, a next-generation model that promises improved reasoning and lower latency. Anthropic redeployed Claude Fable 5, which brings enhanced safety features and better performance on code generation tasks. Both are accessible via Spring AI's provider abstraction, so you can swap them in with minimal code changes.

Resources for Java Developers

If you're looking to get started, the Spring AI Recipes site (habuma.com/springairecipes) has practical examples for common patterns like RAG, function calling, and multi-agent coordination. The podcast episode is a good companion to the official Spring AI 2.0 GA announcement from June 2026.

Bottom line: Spring AI 2.0 is a solid step forward for Java developers who want to build AI-powered applications without fighting provider-specific SDKs. The unified tool calling model alone is worth the upgrade if you're doing anything beyond simple chat completions.