Google Cloud introduces Agent Compute Engine for AI agents

Oct 23, 2026
Google Cloud introduces Agent Compute Engine, a compute service that gives AI agents isolated microVMs for running code, tools and workflows. A preview will begin in the coming months.

Google Cloud announced in a company blog post Agent Compute Engine, a compute service that gives each AI agent an isolated microVM for running code, tools, browsers and workflows. The service already supports Gemini agent tasks and will enter preview in the coming months.

Agent Compute Engine can pause idle microVMs and resume them from snapshots when an agent needs to continue. Each microVM receives persistent Hyperdisk storage, a verified identity and short lived credentials. Google Cloud uses its Titanium infrastructure and a hardware isolated Cloud Hypervisor boundary to separate workloads.

Google Cloud also introduced Agent Substrate, an open source execution runtime for Kubernetes. The company says tests involving 200,000 agents sustained 500 activations per second, with suspend and resume latency below 500 milliseconds. The runtime supports microVMs or gVisor and can run with agent frameworks including Claude Code, OpenClaw and Hermes.

The company is also extending Gemini agent governance across Google Cloud, other cloud platforms, local infrastructure and managed devices. Agent Gateway will add granular cost controls, model routing and simpler custom integrations, with previews due this quarter.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.

Also, consider following us on social media:

AI Chips & Datacenters AI Brief AI Brief (X)
Free newsletter

Silicon Brief

Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.

AI Token Webinar

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You may also like

Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.