Agent Platform Samples

A curated list of code samples and tutorials for building agents on Gemini Enterprise Agent Platform (previously Vertex AI).

Watch the Intro Video

🛠️ Build - Turning code into agents

• Agent Development Kit (ADK) - Open-source framework for building agents.
  • Agents CLI - Command-line tool for agent development.
• Agent Studio - Low-code visual agent builder.
• Agent Garden - Discover and deploy pre-built agents.
• Models - Foundation models powering agent intelligence.
  • Gemini API - Access Google's foundation models.
  • Model Garden - Discover foundation and open-weight models.
  • Model Inference - Enterprise-grade model serving infrastructure.
  • Managed Training - Fully managed model fine-tuning.
• Tools, data, and other agents - Connect and extend agent capabilities.
  • Agent2Agent Protocol (A2A) - Secure multi-agent coordination standard.
  • Grounding - Connect models to real-world data.
  • RAG Engine - Retrieve relevant enterprise information.
  • Model Context Protocol (MCP) - Standard protocol for external tools.
  • Vector Search - Managed vector embedding database.
  • Agent Search (previously Vertex AI Search) - Custom search engines for your own data.
  • Agent-to-UI (A2UI) - Dynamic user interface generation protocol.
  • Agent Payments Protocol (AP2) - Secure automated payment execution protocol.
  • Universal Commerce Protocol (UCP) - Standard for managing e-commerce operations.

📈 Scale - Reliability over time

• Agent Runtime (previously Agent Engine) - Managed platform for serving agents.
• Agent Sessions - Track interactions across multiple conversations.
• Agent Memory Bank - Persistent long-term agent memory.
• Agent Sandbox - Secure environment for code execution and computer use.

🔐 Govern - Trust as a feature

• Agent Gateway - Central control, authentication, and rate-limiting. (Private Preview)
• Agent Identity - Unique credentials for agents.
• Agent Policies - IAM Policies for agents. (Private Preview)
• Agent Registry - Single source of approved tools and agents.
• Model Armor - Prevents prompt injections and leaks.

⚙️ Optimize - Evaluation and enhancement

• Agent Evaluation - Automatically test complex agent interactions.
• Agent Simulation - Generate synthetic interactions for stress-testing.
• Agent Observability - Forensic visibility into agent reasoning.
• Agent Optimizer - Automate improvements based on failures.

Developer's guides

• Onboarding Guide - A quick-start guide accelerates your onboarding to Agent Platform

🪜 Other Tutorials & Notebooks

• Gemini
  • Intro to Gemini 3.1 Pro
  • Intro to Gemini 3 Flash
  • Intro to Gemini 3.1 Flash-Lite
  • Intro to Agent Runtime
  • Intro to Batch Prediction
  • Intro to Code Execution
  • Intro to Agentic Vision
  • Intro to Computer Use
  • Intro to Context Caching
  • Intro to Evaluation
  • Intro to Grounding
  • Intro to Live API
  • Intro to LangGraph
  • Intro to Prompt Attacks and Mitigation
  • Intro to Supervised Fine Tuning
  • Intro to Document Processing
  • Intro to Spatial Understanding
  • YouTube Video Analysis
• Image Generation (Nano Banana)
  • Gemini 3.1 Flash Image (Nano Banana 2 🍌)
  • Gemini 3 Pro Image (Nano Banana Pro 🍌)
• Veo
  • Veo 3.1 Video Generation
  • Veo 3.1 Advanced Controls
• Lyria
  • Lyria 3 Music Generation
• Embeddings
  • Getting Started with Text Embeddings + Vector Search
  • Introduction to Multimodal Embeddings
  • Introduction to Vertex AI Vector Search 2.0
  • Combining Semantic & Keyword Search: A Hybrid Search Tutorial
• Model Garden
  • Intro to Model Garden

---

This is not an officially supported Google product. This project is not eligible for the Google Open Source Software Vulnerability Rewards Program.