Google Announces General Availability of Interactions API as Primary Gemini Interface

Google has made a significant move in its AI developer ecosystem by announcing the general availability of the Interactions API, positioning it as the go-to interface for interacting with Gemini models and agents. This API, which was initially launched in public beta in December 2025, has quickly become a favored tool among developers for building applications powered by Gemini. The Interactions API provides a single, unified endpoint that simplifies the process of building with Gemini, whether developers are calling a model for inference or running an agent for autonomous tasks. It supports server-side state, background execution for long-running processes, tool combination, and multimodal generation, making it versatile for complex AI applications. According to Google, the API is designed from the ground up to handle stateful, agentic workflows, which are becoming increasingly crucial for advanced AI development. To support this, a `gemini-interactions-api Skill` has been introduced to inject best-practice patterns for Interactions API development directly into agents' contexts, covering aspects like streaming, function calling, structured output, and deep research. This new standard is now the default for Google AI Studio, the Gemini API, and all associated documentation. While the legacy `generateContent` API will continue to be fully supported and receive mainline Gemini models, Google strongly recommends using the Interactions API for all new projects. The company anticipates that cutting-edge capabilities for long-running models and agents will primarily be rolled out exclusively on the Interactions API, underscoring its strategic importance for the future of Gemini development.