Gemini Cloud Assist serves as your AI-powered collaborator, helping you move from reactive troubleshooting to proactive, autonomous cloud operations. By integrating deeply into the cloud platform, Gemini Cloud Assist simplifies the end-to-end application lifecycle, offering intelligent, agentic guidance for designing, deploying, troubleshooting, and optimizing your workloads.
Gemini Cloud Assist is a multi-agent ecosystem capable of dynamic reasoning, executing iterative tool calls, and maintaining long-term memory across complex tasks. It not only understands your cloud infrastructure but also your application code, organizational policies, and real-time operational state.
This document shows the types of generative AI assistance that are available to all Google Cloud users and available to Gemini Code Assist Enterprise users. Learn about Gemini Cloud Assist pricing .
The Gemini large language models (LLMs) that are used by Gemini Code Assist are trained on datasets of publicly available code, Google Cloud-specific material, and other relevant technical information in addition to the datasets used to train the Gemini foundation models . Models are trained so that Gemini Code Assist responses are as useful to Gemini Code Assist users as possible.
Additionally, Gemini Cloud Assist uses the following details as context when generating responses and assistance:
- Your Google Cloud project ID and organization ID.
- Page context awareness: The current Google Cloud product page you are viewing in the Google Cloud console, including the page URL and all visible text on the screen. This helps ensure responses are highly contextualized and helps eliminate hallucinations.
- Memory: Gemini Cloud Assist leverages memory to store session histories, prioritize key details, and maintain context across complex tasks and multiple turns.
Learn how and when Gemini for Google Cloud uses your data .
AI assistance with Gemini Cloud Assist
The following sections describe the types of generative AI assistance that Gemini Cloud Assist offers.
Design and build
The following table shows the types of generative AI assistance for building and designing applications:
| Feature | Available to all Google Cloud users | Code Assist Enterprise |
|---|---|---|
|
Natural language design
.
Design, configure, provision, and manage complex cloud architectures using
natural language interaction within the Application Design Center.
|
||
|
Infrastructure as code (IaC)
.
Iteratively generate or refine Google Cloud CLI commands, `kubectl` manifests,
and Terraform code for infrastructure and organizational policies through
interactive conversations in App Design Center and
Gemini Cloud Assist chat, and simulate the generated policies
directly within chat.
|
||
|
Safety-first resource mutation and authorization.
Before executing potentially impactful commands (such as Google Cloud CLI and
`kubectl`) that modify state, Gemini Cloud Assist bridges the
gap between diagnosis and resolution by generating executable multi-step
remediation plans. While read-only tools run automatically,
Gemini Cloud Assist enforce human-in-the-loop confirmation
for mutations. This safety-first approach ensures that all mutations are
performed using your identity, require explicit consent, and are executed
and recorded in Cloud Audit Logs. To use this capability, an administrator
must first
[Enable Gemini Cloud Assist to modify resources](/cloud-assist/enable-resource-modification).
You interact with this feature in the
[**Cloud Assist** panel](/cloud-assist/chat-panel).
|
||
|
Identity and Access Management (IAM)
. Simplify
IAM tasks by selecting roles, understanding policies, or
generating them using natural language or guided flows.
|
||
|
Google Cloud console chat
.
Get answers to general questions, best practices, and basic how-to
information for Firebase products and services in chat, using project
context.
|
Support and troubleshoot
The following table shows the types of generative AI assistance for diagnosing and resolving issues:
| Feature | Available to all Google Cloud users | Code Assist Enterprise |
|---|---|---|
|
Investigations
.
Diagnose complex issues by creating and managing AI-driven investigations
based on logs, errors, or natural language prompts. Acting as a user-guided
site reliability engineer (SRE), Gemini Cloud Assist can ingest massive amounts of logs, metrics, and configurations
to explore multiple hypotheses in parallel and trace root causes across
multi-tier topologies.
|
||
|
Proactive alert investigations.
Configure Cloud Alerting policies to automatically trigger
Gemini Cloud Assist. When an alert is triggered,
Gemini Cloud Assist analyzes the issue in the background,
clusters related alerts, and generates a preliminary root-cause analysis
with recommended next steps.
|
||
|
Support handoff
.
Escalate investigations or complex issues by connecting directly with
Google Cloud support and providing investigation results as context
in the support case.
|
If enrolled in a support package | |
|
AI-powered support. As the primary entry point for
support, Gemini Cloud Assist streamlines the support
experience by summarizing environment states and error logs for rich context.
Gemini Cloud Assist integrates AI investigation tools (incorporating AI features previously
available through Concierge) to help you resolve issues before escalating
to a human agent. When automated capabilities or quotas are reached, the agent
seamlessly passes the issue to human support engineers. Fields are
automatically filled with information from the chat conversation.
|
If enrolled in a support package | |
|
Database and analytics troubleshooting.
AI-powered troubleshooting workflows to empower developers and simplify the
resolution of complex performance issues across multiple data services, including Cloud SQL for PostgreSQL
, Cloud SQL for MySQL
,
AlloyDB for PostgreSQL, BigQuery, Managed Service for Apache Spark, and Spanner. Gemini Cloud Assist helps users identify slow SQL queries,
manage slot reservations and capacity, troubleshoot performance bottlenecks, and
configure data environments.
|
||
|
Application code troubleshooting. Investigations go
beyond infrastructure. For services like Cloud Run,
Gemini Cloud Assist can trace root causes from infrastructure
signals directly down to application code, identifying issues
across both code and infrastructure.
|
||
|
Artifact Analysis
.
Ask questions about vulnerabilities identified by Artifact Analysis.
|
||
|
Artifact Registry
.
List your Artifact Registry repositories and container images stored in
Docker-format repositories.
|
||
|
Security worker. A proactive agent that helps
non-security professionals triage and remediate security findings from
Security Command Center by translating complex security risks into clear, actionable steps
to resolve external resource exposure, help secure service accounts, and improve overall compliance.
|
||
|
Developer Connect insights (
Preview
).
Learn more about your software development lifecycle and diagnose problems
with your applications.
|
||
|
Google Cloud Observability
. Monitor
system health and troubleshoot issues by exploring metrics, logs, alerts,
and errors using natural language within Google Cloud Observability.
|
||
|
Firebase troubleshooting assistance
.
Troubleshoot specific Firebase product issues using AI-driven expertise and
guidance tailored to project context.
|
Optimize
The following table shows the types of generative AI assistance for optimizing cost and usage:
| Feature | Available to all Google Cloud users | Code Assist Enterprise |
|---|---|---|
|
Cost optimization for applications, SKUs, products,
resources, and workloads.
Ask natural language questions in the chat panel to get precise answers
about your cloud spend. This feature extends Cloud Hub Optimization
and Cost Explorer
.
|
||
|
Cost optimization in FinOps Hub
.
Optimize cloud spend by analyzing costs, trends, waste, and utilization with
tailored FinOps views. Also includes AI-drafted summaries for collaboration.
|
||
|
Proactive cost anomaly analysis.
Detected cost anomalies can proactively trigger an analysis to explain the
root cause of a cost change. This analysis runs once a day when an anomaly
is detected in a project. The agent analyzes the cost spike, analyzes
products, SKUs, and resources to identify the infrastructure changes that
caused the cost spike along with the user / service account responsible for
those changes. When applicable, the agent proposes potential strategies to
fix such cost spikes. Gemini can make mistakes, so double-check any
proposed strategies before implementing them.
|
||
|
Database optimization
.
AI-driven insights to optimize database costs and potentially achieve savings by
configuring instances based on usage patterns.
|
||
|
Flow analyzer
.
Analyze network flows and configurations using natural language queries
within the Network Intelligence Center.
|
||
|
Storage Insights
.
Understand and manage storage services by querying usage, configuration, and
insights using natural language.
|
Core platform capabilities
The following table shows the core platform capabilities of Gemini Cloud Assist:
| Feature | Available to all Google Cloud users | Code Assist Enterprise |
|---|---|---|
|
Agentic reasoning. Gemini Cloud Assist
employs dynamic reasoning loops, iteratively invoking tools (such as Cloud APIs
and Kubernetes APIs) to gather data, evaluate state, and determine the next
best action.
|
||
|
User-guided multiturn agent. Provides multi-turn,
user-guided chat interactions. You can ask for clarification, guide the agent's
focus, receive clarifying questions, and dynamically change the scope of your
request without losing context.
|
||
|
Agent identity. When operating proactively in the background,
Gemini Cloud Assist utilizes a dedicated Service Account
(Agent Identity) rather than your personal credentials. This enforces the
principle of least privilege and ensures all autonomous actions are fully
audit-logged.
|
||
|
MCP support. Extend Gemini Cloud Assist
anywhere using the Model Context Protocol (MCP). You can invoke agents directly
from third-party client agents, local integrated development environments (IDEs)
like Gemini CLI, or Claude Code.
|
||
|
Custom instructions. Administrators and users can provide
persistent, custom system instructions. This lets you define specific agent
personas (for example, a strict SRE), enforce response formats, or provide
supplemental business context like naming conventions.
|
Where to interact with Gemini Cloud Assist
The following table describes where to interact with Gemini Cloud Assist.
| Where to interact with Gemini Cloud Assist | Available to all Google Cloud users | Code Assist Enterprise |
|---|---|---|
|
Cloud Assist panel
. The Cloud Assistpanel lets you
get conversational assistance in the Google Cloud console and the
Cloud Mobile App.
|
||
|
Google Cloud console quick prompts. Various buttons and
interactive elements are available in the Google Cloud console and
Cloud Mobile App.
|
||
|
Cloud Hub
. A
hub available in the Google Cloud console that empowers DevOps teams to optimize
resource and application costs and utilization.
|
||
|
Database Center
.
Simplifies development with efficient queries and optimized schemas. It
enhances security by identifying vulnerabilities, automates migrations, and
provides centralized management for database fleets, boosting performance
and reducing costs.
|
||
|
Application Design Center
.
Design optimal cloud applications that can be managed throughout their
lifecycle with Gemini in the App Design Center.
Describe your technical and business requirements to Gemini
in the App Design Center, and it collaboratively produces
solutions for consideration, refinement, and deployment.
|
||
|
Investigations
.
The Gemini Cloud Assist investigations feature is an
AI-enabled tool for analyzing, troubleshooting, and diagnosing technical
issues, available within the Google Cloud console, Cloud Mobile App app,
command-line interface, and IDEs. For example, you can run an investigation
if you receive an error message or are unable to accomplish a task because
of unexpected behavior in the Google Cloud console.
|
||
|
Third-party integrations.
Integrate Gemini Cloud Assist with third-party developer
productivity tools such as Slack, WhatsApp, and ServiceNow.
|
Set up Gemini Cloud Assist
For detailed setup steps, see Set up Gemini Cloud Assist .
What's next
- See the latest enhancements and fixes in release notes .
- Learn about Gemini Cloud Assist pricing .
- Learn how to use the Cloud Assistpanel in the Google Cloud console .
- Learn how to set up Gemini Cloud Assist .
- Learn how Cloud Storage for Google Cloud uses your data .
