Examples and Samples

Genkit provides a comprehensive collection of examples and sample applications to help you get started and learn different features of the framework.

Interactive Examples

Play with AI sample apps with visualizations of the Genkit code that powers them, at no cost to you.

Genkit by Example

Explore interactive examples in your browser

Sample Applications by Language

JavaScript/TypeScript Samples

The JavaScript samples demonstrate various Genkit features and use cases:

Basic Samples

Coffee Shop

Simple “AI barista” demonstrating basic LLM usage

Gemini Basics

Getting started with Google’s Gemini models

OpenAI Integration

Using OpenAI models with Genkit

Prompting Techniques

Various prompting patterns and techniques

Intermediate Samples

Menu Understanding

Progressive restaurant menu app with 5 iterations:

Basic prompts with templates
Flows with strong schemas and tools
Multi-turn chat with session history
Vector database and RAG
Vision model for photo menu ingestion

Character Generator

RPG character generation with structured output

Chatbot

Full-stack chatbot with JavaScript frontend and multi-model support

Angular Streaming

Angular application with streaming AI responses

Advanced Samples

School Agent System

Multi-agent architecture with routing:

RoutingAgent: Main entry point and router
AttendanceAgent: Specialized for absence/tardy reporting
GradesAgent: Manages academic performance queries

Demonstrates agent handoff, specialized tools, and separation of concerns

Go Samples

The Go SDK includes extensive samples covering various capabilities:

Basic Go Samples

basic: Simple text generation
basic-prompts: Working with prompt templates
basic-structured: Structured output generation
anthropic: Using Anthropic Claude models
coffee-shop: AI barista example in Go

Model Integration Samples

gemini: Google Gemini integration
cache-gemini: Context caching with Gemini
code-execution-gemini: Server-side code execution
imagen: Image generation with Imagen
imagen-gemini: Combined image and text generation
compat_oai: OpenAI-compatible API integration (Anthropic, Custom, OpenAI)

Advanced Go Samples

flow-sample1: Building flows
formats: Output formatting examples
intermediate-interrupts: Human-in-the-loop workflows
durable-streaming: Streaming with durability
durable-streaming-firestore: Streaming with Firestore backend
firebase-retrievers: Firebase vector search integration
files-api-vision: Vision with Files API
menu: Restaurant menu understanding (Go port)

MCP (Model Context Protocol) Samples

mcp-client: MCP client implementation
mcp-server: MCP server with client example
mcp-ception: Advanced MCP patterns
mcp-git-pr-explainer: GitHub PR analysis using MCP

View All Go Samples

Browse the complete Go samples directory

Python Samples (Alpha)

Python samples are organized by category with consistent structure:

Model Provider Samples

Google AI & Vertex AI

provider-google-genai-hello: Gemini with Google AI
provider-google-genai-vertexai-hello: Gemini with Vertex AI
provider-google-genai-code-execution: Server-side code execution
provider-google-genai-context-caching: Context caching for long prompts
provider-google-genai-vertexai-image: Imagen image generation
provider-google-genai-media-models-demo: TTS, Veo, Lyria, Imagen, editing

Third-Party Providers

provider-anthropic-hello: Claude models
provider-amazon-bedrock-hello: AWS Bedrock (Claude, Llama, Nova)
provider-microsoft-foundry-hello: Azure AI Foundry + Application Insights
provider-ollama-hello: Local models with Ollama
provider-compat-oai-hello: OpenAI-compatible APIs
provider-deepseek-hello: DeepSeek V3 and R1
provider-xai-hello: Grok models
provider-cloudflare-workers-ai-hello: Cloudflare Workers AI
provider-mistral-hello: Mistral models
provider-huggingface-hello: HuggingFace Inference API
provider-cohere-hello: Cohere models and embeddings

Specialized Features

provider-vertex-ai-model-garden: Third-party models via Model Garden
provider-vertex-ai-rerank-eval: Rerankers and evaluators
provider-vertex-ai-vector-search-bigquery: BigQuery with vectors
provider-vertex-ai-vector-search-firestore: Firestore with vectors
provider-firestore-retriever: Firestore vector search
provider-checks-hello: Google Checks AI safety guardrails
provider-observability-hello: Sentry, Honeycomb, Datadog, Grafana, Axiom

Framework Feature Samples

framework-context-demo: Context propagation through flows
framework-dynamic-tools-demo: Dynamic tool registration
framework-evaluator-demo: Custom evaluators and RAGAS
framework-format-demo: Output formatting and schemas
framework-middleware-demo: Custom retry and logging middleware
framework-prompt-demo: Dotprompt features (picoschema, roles, media, variants, helpers)
framework-realtime-tracing-demo: Real-time tracing visualization
framework-restaurant-demo: Restaurant ordering with tools and RAG
framework-tool-interrupts: Human-in-the-loop tool approval

Web Framework Samples

web-endpoints-hello: FastAPI, Litestar, Quart, gRPC endpoints
web-fastapi-bugbot: AI code reviewer with streaming SSE
web-flask-hello: Flask HTTP endpoints
web-multi-server: Multiple Genkit servers (Litestar, Starlette)
web-short-n-long: ASGI deployment with long-running flows

View All Python Samples

Browse the complete Python samples directory with detailed README

Running Samples

JavaScript/TypeScript

cd samples/<sample-name>
npm install
export GEMINI_API_KEY=your_api_key  # or other provider key
npm run genkit:dev

Access the Developer UI at http://localhost:4000

Go

cd go/samples/<sample-name>
go mod tidy
export GOOGLE_GENAI_API_KEY=your_api_key
genkit start -- go run .

Access the Developer UI at http://localhost:4000

Python

cd py/samples/<sample-name>
./run.sh  # Starts DevUI and runs with hot reload

Each Python sample includes a run.sh script with automatic dependency installation and hot reloading.

Try in Cloud IDEs

IDX (Google)

Try JavaScript samples in Google’s Cloud-Based IDE:

Firebase Studio

Skip local setup entirely and try Genkit using Firebase Studio:

Sample Categories

By Complexity

Basic: Simple single-feature demonstrations (coffee-shop, basic prompts, simple generation)
Intermediate: Multi-feature integration (menu app, chatbot, character generator, streaming)
Advanced: Complex architectures (multi-agent systems, RAG pipelines, MCP integration)

By Feature

Text Generation: coffee-shop, gemini, openai, prompts
Structured Output: character-generator, basic-structured, formats
Tool Calling: menu (step 2+), schoolAgent, framework-dynamic-tools-demo
Chat/Sessions: chatbot, angular, framework-restaurant-demo
RAG: menu (step 4), framework-restaurant-demo
Multimodal: menu (step 5), files-api-vision, imagen samples, media-models-demo
Streaming: angular, chatbot, durable-streaming, fastapi-bugbot
Multi-Agent: schoolAgent
Evaluation: framework-evaluator-demo, vertex-ai-rerank-eval

Contributing Examples

We welcome new examples! See our Contributing Guide for guidelines on:

Example structure and naming conventions
Documentation requirements
Testing and validation
Submitting pull requests

Additional Resources

Community Examples

Community-contributed samples and examples

Documentation

Official Genkit documentation

GitHub Discussions

Ask questions and share your examples

Discord Community

Get help and connect with other developers

Community

​Interactive Examples

Genkit by Example

​Sample Applications by Language

​JavaScript/TypeScript Samples

​Basic Samples

Coffee Shop

Gemini Basics

OpenAI Integration

Prompting Techniques

​Intermediate Samples

Menu Understanding

Character Generator

Chatbot

Angular Streaming

​Advanced Samples

School Agent System

​Go Samples

​Basic Go Samples

​Model Integration Samples

​Advanced Go Samples

​MCP (Model Context Protocol) Samples

View All Go Samples

​Python Samples (Alpha)

​Model Provider Samples

​Framework Feature Samples

​Web Framework Samples