Skip to main content

​
Plugins Overview

Genkit plugins extend the framework with additional capabilities like model providers, vector stores, telemetry, and integrations. Plugins provide a standardized way to add new models, embedders, retrievers, indexers, and other components to your Genkit application.

​
What are Plugins?

Plugins in Genkit are modular extensions that:
  • Provide models from various AI providers (Google AI, Anthropic, OpenAI, etc.)
  • Add vector stores for retrieval-augmented generation (RAG)
  • Enable telemetry and monitoring capabilities
  • Integrate frameworks like Express, Firebase, and Next.js
  • Extend functionality with custom embedders, evaluators, and more

​
Plugin Categories

​
Model Providers

Model provider plugins give you access to AI models from different vendors:

​
Vector Stores

Vector store plugins enable RAG with various vector databases:
  • genkitx-chroma - ChromaDB vector store
  • genkitx-pinecone - Pinecone vector database
  • @genkit-ai/dev-local-vectorstore - Local development vector store
  • @genkit-ai/cloud-sql-pg - Google Cloud SQL PostgreSQL with pgvector

​
Framework Integrations

Integration plugins connect Genkit with popular frameworks:

​
Telemetry & Monitoring

  • @genkit-ai/firebase - Firebase telemetry and monitoring
  • @genkit-ai/google-cloud - Google Cloud operations

​
Installation and Usage

​
Installing Plugins

Install plugins using your package manager: 
npm install @genkit-ai/google-genai
npm install @genkit-ai/anthropic
npm install genkitx-ollama

​
Basic Configuration

Plugins are configured when initializing Genkit: 
import { genkit } from 'genkit';
import { googleAI } from '@genkit-ai/google-genai';
import { anthropic } from '@genkit-ai/anthropic';

const ai = genkit({
  plugins: [
    googleAI(),
    anthropic({ apiKey: process.env.ANTHROPIC_API_KEY }),
  ],
});

​
Using Plugin Resources

Once configured, use models and other resources from plugins: 
// Using a Google AI model
const response = await ai.generate({
  model: googleAI.model('gemini-2.5-flash'),
  prompt: 'Tell me about Genkit plugins',
});

// Using an Anthropic model
const response2 = await ai.generate({
  model: anthropic.model('claude-sonnet-4-5'),
  prompt: 'Explain AI plugins',
});

​
Plugin API

Plugins implement the GenkitPlugin interface: 
export interface PluginProvider {
  name: string;
  initializer: () =>
    | InitializedPlugin
    | void
    | Promise<InitializedPlugin | void>;
  resolver?: (action: ActionType, target: string) => Promise<void>;
  listActions?: () => Promise<ActionMetadata[]>;
}

export interface InitializedPlugin {
  models?: Action<z.ZodTypeAny, z.ZodTypeAny>[];
  retrievers?: Action<z.ZodTypeAny, z.ZodTypeAny>[];
  embedders?: Action<z.ZodTypeAny, z.ZodTypeAny>[];
  indexers?: Action<z.ZodTypeAny, z.ZodTypeAny>[];
  evaluators?: Action<z.ZodTypeAny, z.ZodTypeAny>[];
}

​
Official Plugins

​
Model Providers

  • @genkit-ai/google-genai - Google AI and Vertex AI (Gemini, Imagen, Lyria)
  • @genkit-ai/anthropic - Anthropic Claude models
  • genkitx-ollama - Local Ollama models
  • @genkit-ai/compat-oai - OpenAI-compatible models

​
Vector Stores

  • genkitx-chroma - ChromaDB
  • genkitx-pinecone - Pinecone
  • @genkit-ai/dev-local-vectorstore - Local development
  • @genkit-ai/cloud-sql-pg - Cloud SQL PostgreSQL

​
Integrations

  • @genkit-ai/express - Express.js
  • @genkit-ai/firebase - Firebase
  • @genkit-ai/next - Next.js
  • @genkit-ai/google-cloud - Google Cloud

​
Other

  • @genkit-ai/evaluators - AI evaluation tools
  • @genkit-ai/checks - Quality checks
  • @genkit-ai/langchain - LangChain integration
  • @genkit-ai/mcp - Model Context Protocol

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us