Cloud Run AI Cookbook

Hands-on with Gemma 3 on Google Cloud

This blog post announces two codelabs that show developers how to deploy Gemma 3 on Google Cloud using either Cloud Run for a serverless approach or Google Kubernetes Engine (GKE) for a platform approach.

Easy AI workflow automation: Deploy n8n on Cloud Run

This blog post explains how to deploy agents using the n8n workflow automation tool on Cloud Run to create AI-powered workflows and integrate with tools like Google Workspace.

Automate app deployment and security analysis with new Gemini CLI extensions

This blog post announces the Cloud Run extension in the Gemini CLI to simplify application deployment with a single /deploy command.

From localhost to launch: Simplify AI app deployment with Cloud Run and Docker Compose

This blog post announces a collaboration between Google Cloud and Docker that simplifies the deployment of complex AI applications by allowing developers to use the gcloud run compose up command to deploy their compose.yaml files directly to Cloud Run.

Build and Deploy a Remote MCP Server to Google Cloud Run in Under 10 Minutes

This blog post provides a step-by-step guide to building and deploying a secure, remote Model Context Protocol (MCP) server on Google Cloud Run in under 10 minutes using FastMCP, and then testing it from a local client.

AI deployment made easy: Deploy your app to Cloud Run from AI Studio or MCP-compatible AI agents

This blog post introduces ways to simplify AI deployments with one-click deployment from AI Studio to Cloud Run, direct deployment of Gemma 3 models, and a MCP server for agent-based deployments.

50% faster merge and 50% fewer bugs: How CodeRabbit built its AI code review agent with Google Cloud Run

This article showcases how CodeRabbit, an AI code review tool, utilizes Cloud Run to build a scalable and secure platform for executing untrusted code, ultimately cutting code review time and bugs in half.

Create shareable generative AI apps in less than 60 seconds with Vertex AI and Cloud Run

This article introduces a feature in Vertex AI that allows for one-click deployment of web applications on Cloud Run. Use generative AI prompts to streamline the process of turning a generative AI concept into a shareable prototype.

How to deploy serverless AI with Gemma 3 on Cloud Run

This blog post announces Gemma 3, a family of lightweight, open AI models, and explains how to deploy them on Cloud Run for scalable and cost-effective serverless AI applications.

Unlock Inference-as-a-Service with Cloud Run and Vertex AI

This blog post explains how developers can accelerate the development of generative AI applications by adopting an Inference-as-a-Service model on Cloud Run. This enables hosting and scaling of LLMs with GPU support and integrating them with Retrieval-Augmented Generation (RAG) for context-specific responses.

RAG infrastructure for generative AI using Vertex AI and Vector Search

This document presents a reference architecture for building a generative AI application with Retrieval-Augmented Generation (RAG) on Google Cloud, utilizing Vector Search for large-scale similarity matching and Vertex AI for managing embeddings and models.

Stop coding, start architecting: Google Antigravity + Cloud Run

This video introduces Google's agentic IDE, Antigravity. Use it to build and deploy a full stack app to Cloud Run from scratch. Watch this video to write a spec sheet for the AI, force it to use modern Node.js (no build steps!), and watch it autonomously debug a port mismatch during deployment touching a config file.

This AI agent runs on Cloud Run + NVIDIA GPUs

This video shows how to build a real AI agent application on a serverless NVIDIA GPU. See a demo of a smart health agent that uses open source models like Gemma with Ollama on Cloud Run, and LangGraph to build a multi-agent workflow (RAG + tools).

Power your AI agents with MCP tools on Google Cloud Run

This video introduces MCP (Model Context Protocol) and how it makes life easier for AI agent developers. Get a walk through of building an MCP server using FastMCP, and deploying an ADK agent on Cloud Run. See how the code handles service to service authentication using Cloud Run's built-in OIDC tokens.

We tried to jailbreak our AI (and Model Armor stopped it)

This video shows an example of using Google's Model Armor to block threats with an API call.

Don't guess: How to benchmark your AI prompts

This video shows how to use Vertex AI to build reliable generative AI applications using Google Cloud's tools. Developers will learn how to use Google Cloud tools for rapid prototyping, get hard numbers with data-driven benchmarking, and finally, build an automated CI/CD pipeline for true quality control, all while avoiding common pitfalls.

How to build a multi-agent app with ADK and Gemini

This video shows how to build an app using Google's ADK (Agent Development Toolkit) that helps you refine and collaborate on content. Explore how stateful multi-agents work better than a single agent.

Build an AI app that watches videos using Gemini

This video shows how to build an app that watches and understands YouTube videos using Gemini 2.5 Pro. Use smart prompts to customize your app's output for blog posts, summaries, quizzes, and more. This video covers how to integrate Gemini to generate both text content and header images from video input, discuss cost considerations, and explain how to handle longer videos with batch requests.

Let's build a GenAI app on Cloud Run

This video walks you through the architecture and code, using AI to help with every step.

Build AI agents with Cloud Run and Firebase Genkit

This video shows how to build AI agents with Cloud Run and Firebase Genkit, a serverless AI agent builder.

Cloud AI: it's just an API

This videos provides a demo on how to quickly build a tech support application using AI Studio, Cloud Functions, and Firebase Hosting. Learn how to leverage Large Language Models (LLMs) and see a practical example of integrating AI into a traditional web application.

Building AI agents on Google Cloud

This video shows how to build and deploy AI agents using Cloud Run and Vertex AI. Explore key concepts like tool calling, model agnosticism, and the use of frameworks like LangGraph and the Agent Development Kit (ADK).

How to host DeepSeek with Cloud Run GPUs in 3 steps

This video shows how to simplify hosting the DeepSeek AI model with Cloud Run GPUs. Learn how to deploy and manage Large Language Models (LLMs) on Google Cloud with three commands. Watch along and discover the capabilities of Cloud Run and the Ollama command-line tool, allowing developers to operate AI applications rapidly with on-demand resource allocation and scaling.

How to use Gemini function calling with Cloud Run

This video explores the power of Gemini function calling and learn how to integrate external APIs into your AI applications. Build a weather app that leverages Gemini's natural language understanding to process user requests and fetch weather data from an external API, providing a practical example of function calling in action.

Text to image with Google Cloud's Vertex AI on Cloud Run

This video shows how to build an image generation app using Vertex AI on Google Cloud. With Vertex AI image generation model, developers can create stunning visuals without the need for complex infrastructure or model management.

Ollama and Cloud Run with GPUs

This video explains how to use Ollama to easily deploy large language models on Cloud Run with GPUs for scalable and efficient AI model deployment in the cloud.

Protecting sensitive data in AI apps

This video shows how to safeguard sensitive data in AI applications. Explore key concepts, best practices, and tools for protecting data throughout the AI lifecycle.

RAG with LangChain on Google Cloud

This video shows how to enhance the accuracy of your AI applications using Retrieval-Augmented Generation (RAG). Build a web application that leverages the power of RAG with LangChain, a technique that makes AI responses more accurate and precise.

RAG vs Model tuning vs Large prompt window

This video discusses the three primary methods for integrating your data into AI applications: prompts with long context windows, Retrieval Augmented Generation (RAG), and model tuning. Learn the strengths, limitations, and ideal use cases for each approach to make informed decisions for your AI projects in this episode of Serverless Expeditions.

Prompt engineering for developers

This video shows how to use prompt engineering to improve the quality of AI responses. Watch the video to learn how to unlock more accurate and relevant responses from generative AI with chain of thought, few-shot, and multi-shot prompting techniques.

Deploying A GPU-Powered LLM on Cloud Run

This video shows how you can deploy your own GPU-powered large language model (LLM) on Cloud Run. This video walks through taking an open-source model like Gemma and deploying it as a scalable, serverless service with GPU acceleration

Use GPUs in Cloud Run

This video shows a demonstration of deploying Google's Gemma 2, an open-source large language model, through Ollama on Cloud Run.

Build AI chat apps on Google Cloud

This video shows how to build a large language model (LLM) chat app on Gemini.

Multimodal AI

This video shows a demo of using Vertex AI to build a multimodal application that processes video, audio, and text to create output.

Using Serverless Generative AI | Google Vertex AI

This video shows how to build and deploy blazing-fast generative AI apps using Vertex AI Studio, Cloud Run, and generative AI models.

Deploying and Running n8n on Google Cloud Run

This codelab shows how to deploy a production-ready instance of the n8n workflow automation tool on Cloud Run, complete with a Cloud SQL database for persistence and Secret Manager for sensitive data.

How to run LLM inference on Cloud Run GPUs with vLLM and the OpenAI Python SDK

This codelab shows how to deploy Google's Gemma 2 2b instruction-tuned model on Cloud Run with GPUs, using vLLM as an inference engine and the OpenAI Python SDK to perform sentence completion.

Deploy, Manage, and Observe ADK Agent on Cloud Run

This codelab walks you through deploying, managing, and monitoring a powerful agent built with the Agent Development Kit (ADK) on Cloud Run.

How to deploy a secure MCP server on Cloud Run

This codelab walks you through deploying a secure Model Context Protocol (MCP) server on Cloud Run and connecting to it from the Gemini CLI.

Build and deploy an ADK agent that uses an MCP server on Cloud Run

This codelab guides you through building and deploying a tool-using AI agent with the Agent Development Kit (ADK). The agent connects to a remote MCP server for its tools, and is deployed as a container on Cloud Run.

How to fine-tune a LLM using Cloud Run Jobs

This codelab provides a step-by-step guide on how to use Cloud Run Jobs with GPUs to fine-tune a Gemma 3 model on the Text2Emoji dataset and then serve the resulting model on a Cloud Run service with vLLM.

How to run batch inference on Cloud Run jobs

This codelab demonstrates how to use a GPU-powered Cloud Run job to run batch inference on a Llama 3.2-1b model and write the results directly to a Cloud Storage bucket.

Lab 3:Prototype to Production - Deploy Your ADK Agent to Cloud Run with GPU

This codelab demonstrates how to deploy a production-ready Agent Development Kit (ADK) agent with a GPU-accelerated Gemma backend on Cloud Run. The codelab covers deployment, integration, and performance testing.

How to deploy a Gradio frontend app that calls a backend ADK agent, both running on Cloud Run

This codelab demonstrates how to deploy a two-tier application on Cloud Run, consisting of a Gradio frontend and an ADK agent backend, with a focus on implementing secure, authenticated service-to-service communication.

How to deploy a FastAPI chatbot app to Cloud Run using Gemini

This codelab shows you how to deploy a FastAPI chatbot app to Cloud Run.

How to host a LLM in a sidecar for a Cloud Run function

This codelab shows you how to host a gemma3:4b model in a sidecar for a Cloud Run function.

Securely call your Cloud Run service from anywhere

This article provides a Python code example that acquires an identity token to securely call an authenticated Cloud Run service from any environment. The example uses application default credentials (ADC) to authenticate the call.

Serverless AI: EmbeddingGemma with Cloud Run

This article provides a step-by-step guide on how to containerize and deploy the EmbeddingGemma model to Cloud Run with GPUs, and then use it to build a RAG application.

Chain of Trust for AI: Securing MCP Toolbox Architecture on Cloud Run

This article deconstructs a simple hotel booking application built on Google Cloud. It demonstrates a robust, zero-trust security model using service identities, and shows how a secure chain of trust is established from the end-user all the way to the database.

Serverless AI: Qwen3 Embeddings with Cloud Run

This article provides a tutorial on how to deploy the Qwen3 Embedding model to Cloud Run with GPUs. The article also covers containerization with Docker and Ollama, and provides an example of how to use it in a RAG application.

Still Packaging AI Models in Containers? Do This Instead on Cloud Run

This article advocates for a more efficient and scalable architecture for serving large language models (LLMs) on Cloud Run by decoupling model files from the application container, and instead using Cloud Storage FUSE.

Building an AI-Powered Podcast Generator with Gemini and Cloud Run

This article details how to build a serverless AI-powered podcast generator that uses Gemini for content summarization and Cloud Run. The example orchestrates the automated pipeline for generating and delivering daily audio briefings from RSS feeds.

Power your MCP servers with Google Cloud Run

This article explains the purpose of the Model Context Protocol (MCP) and provides a tutorial on how to build and deploy a MCP server on Cloud Run to expose resources as tools for AI applications.

Deploying & Monitoring ML Models with Cloud Run — Lightweight, Scalable, and Cost-Efficient

This article explains how to deploy, monitor, and automatically scale a machine learning model on Cloud Run, utilizing a lightweight monitoring stack with Google Cloud services to track performance and control costs.

Deploying Gemma Directly from AI Studio to Cloud Run

This article provides a step-by-step tutorial on how to take a Gemma model from AI Studio, adapt its code for production, and deploy it as a containerized web application on Cloud Run.

The Triad of Agent Architecture: ADK, MCP, and Cloud Run

This article demonstrates how to build an AI agentic architecture by setting up an Agent Development Kit (ADK) workflow that communicates with a Model Context Protocol (MCP) server hosted on Cloud Run to manage flight bookings.

Exploring Agent2Agent (A2A) Protocol with Purchasing Concierge Use Case on Cloud Run

This article explains the Agent2Agent (A2A) protocol and demonstrates its use with a purchasing concierge application. The Cloud Run app contains multiple AI agents, built with different frameworks, and collaborate amongst itself to fulfill a user's order.

Automating ML Models Deployment with GitHub Actions and Cloud Run

This article provides a comprehensive guide on how to create a CI/CD pipeline with GitHub Actions to automate the build and deployment of machine learning models as containerized services on Cloud Run.

Building Sovereign AI Solutions with Google Cloud - Cloud Run

This article provides a step-by-step guide on how to build and deploy a sovereign AI solution on Google Cloud by using Sovereign Controls by Partners. The examples runs a Gemma model on Cloud Run, ensuring data residency and compliance with European regulations.

From Zero to Deepseek on Cloud Run during my morning commute

This article shows how to rapidly deploy the Deepseek R1 model on Cloud Run with GPUs using Ollama during a morning commute. This article explores advanced topics such as embedding the model in the container, A/B testing with traffic splitting, and adding a web UI with a sidecar container.

How to run (any) open LLM with Ollama on Google Cloud Run [Step-by-step]

This article shows how to host any open LLM, such as Gemma 2, on Google Cloud Run using Ollama. The article also includes instructions for creating a Cloud Storage bucket for model persistence and testing the deployment.

Deployment of Serverless Machine Learning models with GPUs using Google Cloud: Cloud Run

This article provides a step-by-step guide to deploying a machine learning (ML) model with GPU support on Cloud Run. The article covers everything from project setup and containerization to automated deployment with Cloud Build and testing with curl and JavaScript.