The Spring AI engineering team has announced the release of Spring AI 2.0.0-RC1, which is an API stabilization milestone on the road to 2.0.0 GA. This release includes a unified tool execution across all models, with tool execution now handled externally via ChatClient with ToolCallingAdvisor. The internalToolExecutionEnabled property has been removed, and the toolNames API and SpringBeanToolCallbackResolver have been replaced with explicit ToolCallback beans. The ToolCallAdvisor has been renamed to ToolCallingAdvisor, and a new ToolSearchToolCallingAdvisor has been introduced for on-demand tool discovery. Memory advisors are now placed outside ToolCallingAdvisor by default, and the DEFAULT_CHAT_MEMORY_PRECEDENCE_ORDER has been lowered. The release also includes chat memory improvements, such as turn-boundary snapping in MessageWindowChatMemory and avoiding duplicate chat memory in tool prompts. Structured output enhancements have been added, including the ability to access the timestamp field from the spring_ai_chat_memory table programmatically. The release also includes API and code cleanup, model updates, bug fixes, documentation updates, and dependency upgrades. The Spring AI team thanks all the contributors who worked on this release, and resources are available on the project page, GitHub, and Stack Overflow. The release notes and documentation for 2.0.0-RC1 are also available. Overall, this release provides a range of improvements and enhancements to the Spring AI framework, and is an important step towards the final 2.0.0 GA release.