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

# Model Providers Overview

> Choose the right AI model provider for your Genkit application

Genkit supports multiple AI model providers, giving you flexibility to choose the best models for your use case. Each provider offers different models, features, and pricing.

## Supported Providers

Genkit includes official support for the following model providers:

### Google AI Providers

* **[Google AI (Gemini Developer API)](/providers/google-genai)** - Quick access to Gemini models with API key authentication. Ideal for prototyping and smaller projects.
* **[Vertex AI](/providers/vertex-ai)** - Enterprise-grade Google Cloud AI platform with advanced features, IAM integration, and broader model access including Imagen, Lyria, and Model Garden.

### Third-Party Providers

* **[Anthropic](/providers/anthropic)** - Access to Claude models (Haiku, Sonnet, Opus) with advanced reasoning capabilities.
* **[OpenAI-Compatible APIs](/providers/openai-compatible)** - Support for OpenAI, xAI (Grok), DeepSeek, and any OpenAI-compatible endpoint.
* **[Ollama](/providers/ollama)** - Run open-source models locally for privacy and offline usage.

### Custom Providers

* **[Custom Providers](/providers/custom-providers)** - Build your own model provider plugin to integrate any AI service.

## Provider Comparison

<table>
  <thead>
    <tr>
      <th>Provider</th>
      <th>Best For</th>
      <th>Authentication</th>
      <th>Key Features</th>
    </tr>
  </thead>

  <tbody>
    <tr>
      <td>Google AI</td>
      <td>Rapid prototyping, small projects</td>
      <td>API Key</td>
      <td>Gemini models, image/video generation, simple setup</td>
    </tr>

    <tr>
      <td>Vertex AI</td>
      <td>Production apps, enterprise</td>
      <td>GCP IAM / API Key (Express Mode)</td>
      <td>Model Garden, Vector Search, fine-tuning, governance</td>
    </tr>

    <tr>
      <td>Anthropic</td>
      <td>Advanced reasoning tasks</td>
      <td>API Key</td>
      <td>Claude models, extended thinking, document citations</td>
    </tr>

    <tr>
      <td>OpenAI</td>
      <td>GPT models, wide adoption</td>
      <td>API Key</td>
      <td>GPT-4o, o1, DALL-E, Whisper, multi-modal</td>
    </tr>

    <tr>
      <td>Ollama</td>
      <td>Local development, privacy</td>
      <td>None (local)</td>
      <td>Open-source models, offline, no API costs</td>
    </tr>
  </tbody>
</table>

## How to Choose a Provider

### For Prototyping

**Use Google AI** if you want to get started quickly with powerful multimodal models:

```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 quantum computing',
});
```

### For Production

**Use Vertex AI** for enterprise applications with advanced features:

* IAM-based access control
* Integration with other Google Cloud services
* Model versioning and governance
* Access to Model Garden (Anthropic, Meta, and more)
* Vector Search for RAG applications

### For Advanced Reasoning

**Use Anthropic** when you need:

* Extended thinking for complex problem-solving
* Document citations for factual accuracy
* Long context windows (200K+ tokens)
* Prompt caching for efficiency

### For Local Development

**Use Ollama** when you need:

* Privacy (data never leaves your machine)
* Offline capabilities
* No API costs
* Experimentation with open-source models

### For OpenAI Compatibility

**Use OpenAI-Compatible plugin** to connect to:

* OpenAI (GPT-4o, o1, etc.)
* xAI (Grok models)
* DeepSeek
* Any service with an OpenAI-compatible API

## Installation

Install the provider plugin you need:

```bash theme={null}
# Google AI
npm install @genkit-ai/google-genai

# Anthropic
npm install @genkit-ai/anthropic

# OpenAI-compatible
npm install @genkit-ai/compat-oai

# Ollama
npm install genkitx-ollama
```

## Using Multiple Providers

You can configure multiple providers in a single Genkit application:

```typescript theme={null}
import { genkit } from 'genkit';
import { googleAI } from '@genkit-ai/google-genai';
import { anthropic } from '@genkit-ai/anthropic';
import { ollama } from 'genkitx-ollama';

const ai = genkit({
  plugins: [
    googleAI(),
    anthropic({ apiKey: process.env.ANTHROPIC_API_KEY }),
    ollama({
      models: [{ name: 'llama3' }],
      serverAddress: 'http://localhost:11434',
    }),
  ],
});

// Use different models for different tasks
const summary = await ai.generate({
  model: googleAI.model('gemini-2.5-flash'),
  prompt: 'Summarize this text...',
});

const analysis = await ai.generate({
  model: anthropic.model('claude-sonnet-4-5'),
  prompt: 'Analyze this data...',
});

const local = await ai.generate({
  model: ollama.model('llama3'),
  prompt: 'Generate a response...',
});
```

## Next Steps

* [Set up Google AI provider](/providers/google-genai)
* [Set up Vertex AI](/providers/vertex-ai)
* [Set up Anthropic](/providers/anthropic)
* [Configure OpenAI-compatible APIs](/providers/openai-compatible)
* [Run local models with Ollama](/providers/ollama)
* [Build a custom provider](/providers/custom-providers)