Today, Google announced the general availability of the Interactions API, which is now its primary API for interacting with Gemini models and agents. The public beta launched in December 2025. The GA release includes a stable schema and new capabilities such as Managed Agents, background execution, and Gemini Omni (soon). Managed Agents provisions a remote Linux sandbox with a single API call, shipping with the Antigravity agent as default. Background execution is enabled by setting `background=True` on any call. Tool improvements allow mixing built-in tools like Google Search and Google Maps with custom functions, and tool results can return images. Deep Research upgrades include two new agent versions, collaborative planning, native charts, and multimodal grounding. Media generation features image generation with Nano Banana 2, music with Lyria 3, and multi-speaker TTS. The schema has been simplified from Roles to Steps. Cost optimizations include Flex and Priority tiers, with Flex offering a 50% cost reduction, and past interactions have 55-day retention on the paid tier. The Interactions API is now the default for Google AI Studio and the Gemini API.