Before you scale: A guide to Cloud Run cost optimization
Before you scale: A guide to Cloud Run cost optimization
Go to the Optimization Hub → https://goo.gle/4lf5KV3
Mitchell Slep (Engineering Manager, Cloud Run) joins Martin Omander to walk through the practical mechanics of Cloud Run cost optimization. This deep dive moves past basic billing models into technical configurations for tuning active services.
Key technical takeaways:
1️⃣ Preventing overruns: Implementing max instances, using Cloud Armor for rate limiting, and setting up budget alerts to handle unexpected traffic spikes or attacks.
2️⃣ Utilization tuning: Using the Cloud Hub Optimization report to identify services with low CPU and memory utilization—like a bot using only 2% of its allocated resources.
3️⃣ Concurrency settings: Why increasing the concurrency limit (default 80) can reduce your instance count and overall bill without stalling your system.
4️⃣ Committed Use Discounts (CUDs): How to leverage flexible spend commitments across Cloud Run, GKE, and Compute Engine.
Start saving money today!
Watch more Serverless Expeditions → https://goo.gle/ServerlessExpeditions
🔔 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech
#Serverless #GoogleCloud
Speakers: Martin Omander, Mitchell Slep
Products Mentioned: Cloud Armor, Cloud Run, Committed Use Discounts
AI travel planner, powered by Spanner (Multi-model)
The AI travel planner demo showcases how Google Cloud Spanner eliminates database sprawl by consolidating relational (OLTP), graph, vector, and analytical (OLAP) workloads into a single, unified platform. One database, three models: Building a multi-model AI travel planner with Spanner.
Break free from data silos. 🚀 Consolidate your relational, graph, and vector needs into a single, unified platform with Spanner. Like the video, let’s connect, and try Spanner's Multi-model capabilities today: https://goo.gle/4c5CD3P.
Speakers: Neha Bhatnagar, Rakesh Attaluri
Products Mentioned: Spanner, Spanner Graph
Sequential agent pattern
Sequential agent pattern is an AI assembly line, where the output of one agent becomes the input for the next agent in the chain. Amit, a Senior Developer Relations Engineer, shares a use case for this type of AI architecture.
Build your own sequential agent on Google Cloud → https://goo.gle/3NL6YdE
Speaker: Amit Maraj
Products Mentioned: Google Cloud, Agent Development Kit, Cloud Run
When to use generative AI vs. traditional AI vs. no AI
When to use generative AI or traditional AI → https://goo.gle/46cOtoQ
Is AI always right for your needs? And if it is, which type of AI should you use? Join Aja and Jason as they discuss the use cases for when to use generative AI, traditional AI, or no AI. Discover what technique is best for your use case.
Watch more Real Terms for AI → https://goo.gle/AIwordsExplained
🔔 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech
#GoogleCloud
Speaker: Aja Hammerly, Jason Davenport
Products Mentioned: AI Infrastructure
Mastering Cloud Run costs: The 75% rule for choosing a billing model
Pick the right billing model and save money today. → https://goo.gle/4c9NLgd
Are you trying to figure out if Cloud Run is the most cost-effective choice for your next project? Many developers find it difficult to estimate cloud costs without a solid intuition for how billing actually works.
In this episode, Martin Omander sits down with Mitchell Slep, Engineering Manager for Cloud Run, to demystify the two primary billing models and show you exactly how to choose the right one for your architecture.
What you will learn:
1️⃣ Request-based billing: How the default model works, including how to take advantage of the free tier and automatic scaling to zero.
2️⃣ Instance-based billing: Why paying for the full lifetime of an instance can actually save you money for steady workloads or background processing.
3️⃣ The 75% rule: Mitchell shares a professional rule of thumb: if your service is processing requests at least 75% of the time, switching models could lower your bill.
4️⃣ Cost estimation: A step by step look at using the Google Cloud Pricing Calculator to project costs for millions of requests.
Chapters:
0:00 - Intro
1:02 - Two things you need to know
1:31 - Request based billing
3:47 - Instance based billing model
4:29 - The 75% rule
5:09 - Google Cloud pricing calculator
6:00 - Recap
Watch more Serverless Expeditions → https://goo.gle/ServerlessExpeditions
🔔 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech
#Serverless #GoogleCloud
Speakers: Martin Omander, Mitchell Slep
Products Mentioned: Cloud Run, Google Cloud Pricing Calculator
Monitoring configuration and automating detection & remediation for MCP
Securing agent workloads is a continuous process of monitoring, detection, and verification. In this video, we explore how to manage the security posture of your AI agents using AI protection capabilities in Security Command Center (SCC).
Watch along as Aron demonstrates how to maintain a centralized inventory of your AI assets, including agents and MCP servers, and how to utilize Posture Management to detect misconfigurations. You will see how runtime findings from Model Armor, such as jailbreak attempts or indirect prompt injections, are surfaced directly in the SCC dashboard for unified threat management.
We also cover the essential observability tools required for auditing agentic systems. This includes configuring Cloud Logging to capture agent activity. Finally, we discuss defensive strategies, including prioritizing chokepoints and using Sensitive Data Protection (SDP) discovery to identify exposed secrets.
Resources:
Configure agent activity logging → https://goo.gle/46nnjvE
Learn more about AI protection in SCC → https://goo.gle/4r6cF4S
Learn more about defining a security posture in SCC → https://goo.gle/3ZUeJAR
Assess AI security risks in SCC → https://goo.gle/3Zj8zu3
Enforce CMEK encryption for Vertex AI resources → https://goo.gle/4r13faC
Learn more about CMEK → https://goo.gle/4rErD1Z
Secure Credentials for MCP Access with Secret Manager → https://goo.gle/3ZmkHdM
Learn more about toxic combinations and chokepoints → https://goo.gle/4ttLhPZ
Google Secure AI Framework (SAIF) → https://goo.gle/45VWWgi
Speaker: Aron Eidelman
Products Mentioned: Security Command Center, Model Armor, Cloud Logging, Sensitive Data Protection, Customer-managed encryption keys, Vertex AI, Google's Secure AI Framework
Google Cloud Live: Getting started with Antigravity
We’re moving beyond the terminal to explore the Antigravity Editor and Agent Manager. We'll show you how to generate apps with Gemini Flash and make UI modifications just by pointing and clicking.
Join hosts Stephanie Wong and special guests Kevin Hou and Andy Zhang to learn how to get started with Antigravity.
Tune in February 24, 2026 at 9:00 A.M. PST / 12:00 P.M. EST
#Antigravity #Gemini
Speakers: Stephanie Wong, Andy Zhang, Kevin Hou
Products Mentioned: Antigravity, Gemini
Gemini CLI + Google MCPs: Migrate & deploy full stack apps
💻 Try the Codelab → https://goo.gle/4qH2Ady
Stop deploying manually. Watch along and learn how to use the Gemini CLI and Google's new Model Context Protocol (MCP) servers to migrate and deploy a full stack application agentically.
Move beyond simple code generation and see how to chain multiple MCP servers together to act as an intelligent DevOps engineer. This is the future of agentic DevOps on Google Cloud.
Watch along and learn:
1️⃣ Architecting: Using the Developer Knowledge MCP to analyze requirements and identify the best database for our app.
2️⃣ Provisioning & migration: Using the Cloud SQL MCP to create a Cloud SQL instance and migrate our local data automatically.
3️⃣ Deployment: Using Cloud Run to containerize and ship the "AllStrides" full stack app to production.
⏱️ Timestamps:
00:00 - Introduction
00:51 - Configure Google MCPs with Gemini CLI
02:00 - How to list Google MCP tools
03:18 - Use Developer Knowledge MCP server
04:09 - Use Cloud SQL MCP server
05:34 - Use Cloud Run MCP server
🚀 Resources & Links:
📂 Get the code (AllStrides example) → https://goo.gle/3MD64zU
💻 Try the Codelab → https://goo.gle/4qH2Ady
📚 Google MCP documentation → https://goo.gle/46guwxE
📰 Read the launch blog → https://goo.gle/4azuB0B
Speaker: Smitha Kolan
Products Mentioned: Cloud Run, Cloud SQL, Gemini CLI, Model Context Protocol
Intro to Agents: What's new and what we've learned
Join Gemini Enterprise Agent Ready (GEAR) to learn how to build scalable agents. → https://goo.gle/4rTlaQU
Learn how the landscape of AI agents has evolved, from core components like thinking models and the Model Context Protocol (MCP) to architectural patterns like sub-agents and orchestrators. Jason and Aja provide a high level update on what has changed since their last Intro to Agents video and how to define, build, and deploy effective agentic systems today.
Speakers: Aja Hammerly, Jason Davenport
Products Mentioned: Vertex AI Agent Builder, Agent Development Kit
Parallel agent pattern
Parallel agent architecture is set up for speed and efficiency. With this pattern, have your agent do multiple things at once. Amit, a Senior Developer Relation Engineer, shares an example of how a coordinator agent launches multiple agents at the exact same time.
Build a parallel agent system on Google Cloud → https://goo.gle/4bxqBjh
Speaker: Amit Maraj
Products Mentioned: Google Cloud, Agent Development Kit, Cloud Run
Build multimodal AI agents in the Gemini Live Agent Challenge
Learn more and register now → https://goo.gle/RegistrationGeminiLive
Are you ready to move beyond the text box? Welcome to the Gemini Live Agent Challenge! Join our global hackathon and build the future of immersive AI agents on Google Cloud. We're challenging developers to build next generation agents that can help you see, hear, speak, and create using Gemini and Google Cloud.
Important Dates:
Submissions open: February 16, 2026
Submissions close: March 16, 2026 (5:00 P.M. PDT)
#GeminiLiveAgentChallenge #Gemini #Agents
Speakers: Ayo Adedeji, Annie Wang
Products Mentioned: Gemini, Gemini CLI
Google Cloud Live: Unlocking Gemini CLI with Skills, Hooks & Plan Mode
Plan Mode documentation → https://goo.gle/Gemini-CLI-Plan-Mode
Gemini CLI - Introducing Hooks blog → https://goo.gle/Gemini-Cli-Hooks
Last week at Google Cloud Live, we got started with the Gemini CLI, but now it’s time to automate your workflows and make the CLI an active part of your development lifecycle. Join host Greg Baugues and special guest Jack Wotherspoon for Unlocking Gemini CLI with skills, hooks & Plan Mode.
Watch us plan and implement a new app feature (and maybe even plan a surprise party) directly in the terminal.
Learn how to automatically lint your code and avoid writing secrets or passwords to your codebase. Discover how to create a skill to automate the task every developer hates: writing documentation.
Tune in Tuesday, February 17, 2026 at 9:00 A.M. PST / 12:00 P.M. EST
Timestamps:
0:00 - Countdown
1:53 - Intro
3:44 - Livestream overview
4:07 - [App Demo] Memory Wall
6:28 - Hooks
15:11 - Skills
23:21 - How to get started with Skills
31:31 - Skills and Hooks for personal use only
33:35 - Best practices for maintaining and sharing skills with a team
34:39 - Can skills be exported to other CLIs?
37:09 - When to use Plan Mode vs Default Mode?
39:29 - Trying Plan Mode
47:27 - Auto accept edits mode
49:25 - Debugging with MCP playwright server
51:06 - Are skills and extensions the same thing?
52:36 - What is the difference between conductor and plan mode?
54:51 - Chrome web MCP vs Antigravity playwright
58:18 - Deploying changes to app demo
1:01:25 - When to use Antigravity vs. Gemini CLI
1:04:59 - How to get started with Gemini CLI
1:09:53 - Meet the team at Google Cloud Next 2026
#Gemini #GeminiCLI
Speakers: Greg Baugues, Jack Wotherspoon
Products Mentioned: Gemini, Gemini CLI
Error 400: API key not valid. Please pass a valid API key..
Domain code: global
Reason code: badRequest