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

# Installation

> Learn how to install and configure the AI SDK in your project.

# Installation

The AI SDK is available as npm packages. You can install the core SDK and optional provider packages based on your needs.

## Requirements

Before you begin, ensure you have:

* **Node.js**: Version 18 or higher
* **Package manager**: npm, pnpm, yarn, or bun

## Install the core SDK

The `ai` package contains the core SDK functionality including `generateText`, `streamText`, and other core APIs:

<CodeGroup>
  ```bash npm theme={null}
  npm install ai
  ```

  ```bash pnpm theme={null}
  pnpm add ai
  ```

  ```bash yarn theme={null}
  yarn add ai
  ```

  ```bash bun theme={null}
  bun add ai
  ```
</CodeGroup>

## Choose a provider

You have two options for accessing AI models:

### Option 1: Vercel AI Gateway (recommended)

The Vercel AI Gateway provides access to hundreds of models from different providers with a single API key. The gateway provider is included in the `ai` package by default.

<Steps>
  <Step title="Get an API key">
    Sign up for a [Vercel AI Gateway API key](https://vercel.com/d?to=%2F%5Bteam%5D%2F%7E%2Fai%2Fapi-keys%3Futm_source%3Dgateway-models-page%26showCreateKeyModal%3Dtrue\&title=Get+Started+with+Vercel+AI+Gateway)
  </Step>

  <Step title="Add to environment">
    Create a `.env.local` file in your project root:

    ```bash theme={null}
    touch .env.local
    ```

    Add your API key:

    ```env filename=".env.local" theme={null}
    AI_GATEWAY_API_KEY=your_api_key_here
    ```
  </Step>

  <Step title="Start using models">
    You can now access any model using a simple string:

    ```typescript theme={null}
    import { generateText } from 'ai';

    const { text } = await generateText({
      model: 'openai/gpt-4o',
      prompt: 'What is an agent?',
    });
    ```
  </Step>
</Steps>

### Option 2: Direct provider packages

Install specific provider packages to connect directly to AI services:

<CodeGroup>
  ```bash OpenAI theme={null}
  npm install @ai-sdk/openai
  ```

  ```bash Anthropic theme={null}
  npm install @ai-sdk/anthropic
  ```

  ```bash Google theme={null}
  npm install @ai-sdk/google
  ```

  ```bash Mistral theme={null}
  npm install @ai-sdk/mistral
  ```
</CodeGroup>

Then configure your provider credentials:

```env filename=".env.local" theme={null}
OPENAI_API_KEY=your_openai_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here
```

Use the provider in your code:

```typescript theme={null}
import { generateText } from 'ai';
import { openai } from '@ai-sdk/openai';
import { anthropic } from '@ai-sdk/anthropic';

// OpenAI
const result1 = await generateText({
  model: openai('gpt-4o'),
  prompt: 'What is an agent?',
});

// Anthropic
const result2 = await generateText({
  model: anthropic('claude-sonnet-4-5'),
  prompt: 'What is an agent?',
});
```

## Framework-specific packages

If you're building chat or generative UI applications, install the framework integration for your UI library:

<CodeGroup>
  ```bash React theme={null}
  npm install @ai-sdk/react
  ```

  ```bash Vue theme={null}
  npm install @ai-sdk/vue
  ```

  ```bash Svelte theme={null}
  npm install @ai-sdk/svelte
  ```

  ```bash Angular theme={null}
  npm install @ai-sdk/angular
  ```
</CodeGroup>

These packages provide hooks and utilities for building chat interfaces:

```typescript theme={null}
import { useChat } from '@ai-sdk/react';

export default function ChatComponent() {
  const { messages, input, handleInputChange, handleSubmit } = useChat();
  
  // Build your chat UI
}
```

## Additional packages

Depending on your use case, you may need additional packages:

### Schema validation

For structured outputs and tool definitions, install Zod:

```bash theme={null}
npm install zod
```

### Next.js RSC (React Server Components)

For Next.js applications using React Server Components:

```bash theme={null}
npm install ai @ai-sdk/react
```

## Verify installation

Create a simple test file to verify your installation:

```typescript filename="test.ts" theme={null}
import { generateText } from 'ai';
import { openai } from '@ai-sdk/openai';

const result = await generateText({
  model: openai('gpt-4o'),
  prompt: 'Say hello!',
});

console.log(result.text);
```

Run the file:

```bash theme={null}
node --loader ts-node/esm test.ts
```

<Note>
  Make sure you've set up your API keys in the `.env.local` file before running the test.
</Note>

## TypeScript configuration

The AI SDK is written in TypeScript and provides full type safety. Ensure your `tsconfig.json` includes:

```json filename="tsconfig.json" theme={null}
{
  "compilerOptions": {
    "target": "ES2020",
    "module": "ESNext",
    "lib": ["ES2020"],
    "moduleResolution": "bundler",
    "strict": true
  }
}
```

## Next steps

<CardGroup cols={2}>
  <Card title="Quickstart" icon="rocket" href="/quickstart">
    Build your first AI application
  </Card>

  <Card title="Providers" icon="server" href="/providers">
    Learn about available AI providers
  </Card>

  <Card title="AI SDK Core" icon="code" href="/ai-sdk-core">
    Explore core functions and APIs
  </Card>

  <Card title="AI SDK UI" icon="browser" href="/ai-sdk-ui">
    Build chat and generative interfaces
  </Card>
</CardGroup>