Gemini Enterprise Agent Platform

Google's Agents API defines and packages reusable AI agents with a base model, instructions, and a preconfigured environment. Agents run in managed cloud sandboxes rather than on the user's laptop; the environments can be shared with users or teams and accessed through an API call.

Gemini API managed agents are configurable agent harnesses from Google that let an application provision a Linux sandbox with a single API call. The agent can reason, execute code, manage files, and browse the web autonomously; developers can customize it with instructions, skills, data, and external tools, and package reusable agents through the Agents API. Each environment is isolated at the operating-system level and supports persistent state, although inactive environments are eventually deleted and virtual machines can spin down. Managed agents are in public preview and use pay-as-you-go pricing based on model tokens and tool usage.

Google's developer API for building applications with Gemini large language models. It provides access to models like Gemini 2.0 Flash and 2.5 Pro, and also supports Gemma open models. Developers use it for tasks such as custom model testing, retrieval, memory, and tool-calling systems.

Managed Agents is a Google Gemini API service for running the Antigravity agent in persistent remote Linux sandboxes. A call to the Interactions API provisions a sandbox, runs the agent loop, and returns an interaction ID and environment ID; subsequent calls can use the previous interaction ID for conversation context and the environment ID to retain files, installed packages, and other sandbox state. The service supports streaming, file downloads, loadable sources such as Google Cloud Storage and GitHub, reusable skills, secure credential injection, and configurable named agents.

Antigravity agent is a Google DeepMind agent harness for local use and managed remote sandboxes through the Gemini API and Google AI Studio. A single Interactions API call can provision a secure, Google-hosted Linux sandbox identified by an environment ID, where the agent uses Gemini 3.8 Flash to plan, execute Bash, Python, and Node.js code, observe results, and repeat until the task is complete. It can install packages, run tests, build applications, manage files, search Google, fetch web pages, and call custom functions or remote MCP servers. The harness supports reusable skills, configurable tools, multi-turn interactions, streaming, automatic context compaction, persistent sandbox files, synchronous hooks for intercepting and validating code-execution and filesystem operations, and sources loaded from GitHub, Google Cloud Storage, or inline files. Managed environments provide a credential-injecting proxy and named agents, allowing the same workflow to run locally and in the cloud.