Google makes Interactions API the default surface for Gemini agents
Google said its Interactions API is generally available and is now the primary way developers should work with Gemini models and agents, adding managed agents, background execution, richer tool use, media generation, and paid-tier history retrieval.
Read more
Google announced general availability for the Interactions API and positioned it as the primary interface for Gemini models and agents. The release gives the API a stable schema and adds managed agents, background execution, tool-combination improvements, Deep Research upgrades, media generation support, Flex and Priority tiers, and 55-day interaction retention on paid plans. Google says its AI Studio, Gemini API documentation, and ecosystem partners will shift toward Interactions API as the default path for new agentic applications.
Key details: Interactions API reached general availability, Google says documentation now defaults to the API for Gemini models and agents, New capabilities include managed agents, background execution, tool improvements, and media generation.
Why it matters: This is Google turning agent workflows into a formal developer platform, which affects how builders structure state, tools, background jobs, and long-running Gemini applications.