> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/genkit-ai/genkit/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

> Genkit is an open-source framework by Google/Firebase for building production-ready AI-powered applications in JavaScript/TypeScript, Go, and Python.

Genkit gives you a unified set of APIs for integrating AI models, defining type-safe workflows (flows), managing prompts, and building retrieval-augmented generation (RAG) pipelines — all with built-in observability and local developer tooling.

<CardGroup cols={2}>
  <Card title="Quickstart: JavaScript" icon="js" href="/quickstart-js">
    Get up and running with Genkit in Node.js or TypeScript in minutes.
  </Card>

  <Card title="Quickstart: Go" icon="golang" href="/quickstart-go">
    Start building AI features in your Go application with Genkit.
  </Card>

  <Card title="Quickstart: Python" icon="python" href="/quickstart-python">
    Use Genkit's Python SDK (alpha) to add AI capabilities to your Python app.
  </Card>

  <Card title="Core Concepts" icon="book" href="/concepts/flows">
    Learn about flows, models, tools, prompts, and RAG.
  </Card>
</CardGroup>

## What is Genkit?

Genkit simplifies AI application development by providing:

* **A unified model interface** — integrate models from Google Gemini, Anthropic Claude, OpenAI-compatible APIs, Ollama, and more with a consistent API
* **Type-safe flows** — define AI workflows with input/output schema validation using Zod (JS/TS) or Pydantic (Python)
* **Tool calling** — let models invoke your functions automatically with multi-turn reasoning loops
* **Prompt management** — use `.prompt` files (Dotprompt format) with Handlebars templating, YAML frontmatter, and versioning
* **RAG pipelines** — built-in abstractions for embedders, retrievers, and vector stores
* **Developer tooling** — local CLI and Developer UI for testing, tracing, and evaluation
* **Production monitoring** — OpenTelemetry-based tracing exportable to Google Cloud, Firebase, or third-party backends

## Language support

<CardGroup cols={3}>
  <Card title="JavaScript / TypeScript" icon="js">
    **Production-ready.** Full feature support. Published as the `genkit` npm package.
  </Card>

  <Card title="Go" icon="golang">
    **Production-ready.** Full feature support. Available as `github.com/firebase/genkit/go`.
  </Card>

  <Card title="Python" icon="python">
    **Alpha.** Core functionality available. Published as the `genkit` PyPI package.
  </Card>
</CardGroup>

## How it works

<Steps>
  <Step title="Initialize Genkit with plugins">
    Create a Genkit instance and load model provider plugins (Google Gemini, Vertex AI, Ollama, etc.).
  </Step>

  <Step title="Define flows and tools">
    Wrap your AI logic in typed flows. Define tools that models can call automatically.
  </Step>

  <Step title="Test with the Developer UI">
    Use `genkit start` to launch the local Developer UI — run flows, inspect traces, and iterate fast.
  </Step>

  <Step title="Deploy and monitor">
    Deploy to Firebase, Cloud Run, or any platform. Monitor model calls, latency, and errors in production.
  </Step>
</Steps>

## Quick example

<CodeGroup>
  ```typescript TypeScript theme={null}
  import { genkit } from 'genkit';
  import { googleAI } from '@genkit-ai/google-genai';

  const ai = genkit({ plugins: [googleAI()] });

  const { text } = await ai.generate({
    model: googleAI.model('gemini-2.5-flash'),
    prompt: 'Explain what Genkit is in one sentence.',
  });

  console.log(text);
  ```

  ```go Go theme={null}
  package main

  import (
    "context"
    "fmt"
    "log"

    "github.com/firebase/genkit/go/ai"
    "github.com/firebase/genkit/go/genkit"
    "github.com/firebase/genkit/go/plugins/googlegenai"
  )

  func main() {
    ctx := context.Background()
    g := genkit.Init(ctx, genkit.WithPlugins(&googlegenai.GoogleAI{}))

    resp, err := genkit.GenerateText(ctx, g,
      ai.WithModelName("googleai/gemini-2.0-flash"),
      ai.WithPrompt("Explain what Genkit is in one sentence."),
    )
    if err != nil {
      log.Fatal(err)
    }
    fmt.Println(resp)
  }
  ```

  ```python Python theme={null}
  from genkit import Genkit
  from genkit.plugins.google_genai import GoogleAI

  ai = Genkit(plugins=[GoogleAI()], model='googleai/gemini-2.0-flash')

  response = await ai.generate(prompt='Explain what Genkit is in one sentence.')
  print(response.text)
  ```
</CodeGroup>

## Explore further

<CardGroup cols={2}>
  <Card title="Core Concepts" icon="lightbulb" href="/concepts/flows">
    Understand flows, models, tools, and prompts — the building blocks of every Genkit app.
  </Card>

  <Card title="Plugin ecosystem" icon="plug" href="/plugins/overview">
    Browse official plugins for model providers, vector stores, and telemetry backends.
  </Card>

  <Card title="Guides" icon="map" href="/guides/structured-output">
    Practical guides for structured output, streaming, agents, and evaluation.
  </Card>

  <Card title="Deploy your app" icon="cloud" href="/deployment/overview">
    Deploy to Firebase, Cloud Run, or any platform that runs your chosen language.
  </Card>
</CardGroup>