Google upgrades Gemini API agents with Gemini 3.6 Flash and scheduled tasks

Google has updated its Gemini API managed agents with Gemini 3.6 Flash as the default model, adding environment hooks, token budget controls, scheduled execution, sandbox management, and free tier access. The update allows developers to execute custom scripts before or after tool use, set maximum token limits, and create recurring workflows via cron schedules without separate orchestration services.

Google upgrades Gemini API agents with Gemini 3.6 Flash and scheduled tasks The update makes Gemini 3.6 Flash the default model while adding tool call hooks, token limits, scheduled execution, sandbox management, and free tier access. Google has updated https://blog.google/innovation-and-ai/technology/developers-tools/expanding-managed-agents-gemini-api-3-6-flash-hooks/ its Gemini API managed agents with Gemini 3.6 Flash as the default model alongside new controls for security, spending, and automated execution. The Antigravity preview agent now uses Gemini 3.6 Flash automatically, with no code changes required from developers already using the service. Developers can also select another supported model through the agent configuration. Available options include Gemini 3.5 Flash and Gemini 3.5 Flash Lite, which Google positions as its lower cost and lower latency option. Gemini 3.6 Flash is designed for coding, reasoning, multimodal analysis, and repeated tool use. Google says the model completes agentic workflows with fewer reasoning steps, tool calls, and debugging loops than Gemini 3.5 Flash. Managed agents coordinate reasoning, code execution, package installation, file operations, and web retrieval through a single Gemini Interactions API call. Each agent operates inside an isolated cloud sandbox managed by Google. The update introduces environment hooks, allowing developers to execute custom scripts before or after an agent uses a tool. Pre-execution hooks can inspect and block actions such as running code or writing files. Post execution hooks can automatically lint generated code, audit outputs, verify files, or send results to an external endpoint. When a hook rejects a tool call, the action is skipped and the reason is returned to the model, allowing the agent to adjust its approach instead of continuing with the blocked operation. Google said the controls can be used to build security and validation pipelines directly inside the managed sandbox. AI investment banking platform OffDeal is using the feature to verify company logos before its agent places them into presentation materials. Google also added token budget controls for autonomous tasks. Developers can set a maximum total token limit covering input, output, and reasoning consumption. When an agent reaches the limit, execution pauses with an incomplete status rather than continuing to consume tokens. The sandbox state remains available, allowing the task to resume later with a new budget. Managed agents can now run automatically through scheduled triggers. Developers can connect an agent, prompt, environment, and cron schedule to create recurring workflows without operating a separate orchestration service. Each scheduled run can reuse the same sandbox, meaning files and other environment data can remain available between executions. Google’s documentation describes triggers as persistent resources that execute agents automatically according to a defined schedule. A new Environments API also lets developers list, inspect, and delete sandbox sessions. This provides a way to recover environment identifiers after a connection is interrupted or remove sessions before their seven day expiration period. Google is also making managed agents available through free tier Gemini API projects, allowing developers to test agentic workflows without activating billing. The update builds on the May launch of managed agents, which introduced Google’s Antigravity agent, background execution, custom agent configurations, and remote Model Context Protocol server support. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy https://cryptobriefing.com/editorial-policy/ .