Next.js with AI SDK

Learn how to build AI-powered Next.js applications using the AI SDK with the App Router.

Why Next.js?

Next.js is the ideal framework for AI applications:

Server Actions: Call AI functions directly from components
Route Handlers: Create streaming API endpoints
React Server Components: Optimize initial page loads
Edge Runtime: Deploy AI features globally
Streaming: Native support for streaming responses

Prerequisites

Node.js 18+
Basic knowledge of React and Next.js
Vercel AI Gateway API key

Quick Start

Create a new Next.js application:

pnpm create next-app@latest my-ai-app
cd my-ai-app

Install AI SDK packages:

pnpm add ai @ai-sdk/react

Configure your API key:

echo "AI_GATEWAY_API_KEY=your-api-key" > .env.local

Core Patterns

1. Text Generation

Generate text on-demand with a button click. Client Component:

'use client';

import { useState } from 'react';

export default function Page() {
  const [generation, setGeneration] = useState('');
  const [isLoading, setIsLoading] = useState(false);

  const handleGenerate = async () => {
    setIsLoading(true);
    
    const response = await fetch('/api/generate', {
      method: 'POST',
      body: JSON.stringify({
        prompt: 'Why is the sky blue?',
      }),
    });
    
    const data = await response.json();
    setGeneration(data.text);
    setIsLoading(false);
  };

  return (
    <div>
      <button onClick={handleGenerate} disabled={isLoading}>
        {isLoading ? 'Generating...' : 'Generate'}
      </button>
      {generation && <p>{generation}</p>}
    </div>
  );
}

API Route:

import { generateText } from 'ai';

export async function POST(req: Request) {
  const { prompt } = await req.json();

  const { text } = await generateText({
    model: 'openai/gpt-4o',
    system: 'You are a helpful assistant.',
    prompt,
  });

  return Response.json({ text });
}

2. Streaming Chat

Build a real-time chat interface with the useChat hook. Client Component:

'use client';

import { useChat } from '@ai-sdk/react';

export default function Chat() {
  const { messages, input, handleInputChange, handleSubmit } = useChat();

  return (
    <div className="flex flex-col w-full max-w-md py-24 mx-auto">
      {messages.map(m => (
        <div key={m.id} className="whitespace-pre-wrap">
          {m.role === 'user' ? 'User: ' : 'AI: '}
          {m.parts.map((part, i) => {
            if (part.type === 'text') {
              return <span key={i}>{part.text}</span>;
            }
          })}
        </div>
      ))}

      <form onSubmit={handleSubmit}>
        <input
          className="fixed bottom-0 w-full max-w-md p-2 mb-8 border border-gray-300 rounded shadow-xl"
          value={input}
          placeholder="Say something..."
          onChange={handleInputChange}
        />
      </form>
    </div>
  );
}

API Route:

import { streamText, convertToModelMessages, UIMessage } from 'ai';

export const maxDuration = 30;

export async function POST(req: Request) {
  const { messages }: { messages: UIMessage[] } = await req.json();

  const result = streamText({
    model: 'openai/gpt-4o',
    messages: await convertToModelMessages(messages),
  });

  return result.toUIMessageStreamResponse();
}

3. Structured Output

Generate structured data with type safety. Client Component:

'use client';

import { useState } from 'react';

interface Recipe {
  name: string;
  ingredients: string[];
  steps: string[];
}

export default function Page() {
  const [recipe, setRecipe] = useState<Recipe | null>(null);

  const generateRecipe = async () => {
    const response = await fetch('/api/recipe', {
      method: 'POST',
      body: JSON.stringify({ dish: 'chocolate cake' }),
    });
    
    const data = await response.json();
    setRecipe(data.recipe);
  };

  return (
    <div>
      <button onClick={generateRecipe}>Generate Recipe</button>
      {recipe && (
        <div>
          <h2>{recipe.name}</h2>
          <h3>Ingredients:</h3>
          <ul>
            {recipe.ingredients.map((ing, i) => (
              <li key={i}>{ing}</li>
            ))}
          </ul>
          <h3>Steps:</h3>
          <ol>
            {recipe.steps.map((step, i) => (
              <li key={i}>{step}</li>
            ))}
          </ol>
        </div>
      )}
    </div>
  );
}

API Route:

import { generateObject, Output } from 'ai';
import { z } from 'zod';

export async function POST(req: Request) {
  const { dish } = await req.json();

  const { object: recipe } = await generateObject({
    model: 'openai/gpt-4o',
    prompt: `Generate a recipe for ${dish}`,
    output: Output.object({
      schema: z.object({
        name: z.string(),
        ingredients: z.array(z.string()),
        steps: z.array(z.string()),
      }),
    }),
  });

  return Response.json({ recipe });
}

4. Tool Calling

Extend AI capabilities with custom tools.

import { streamText, convertToModelMessages, tool, UIMessage } from 'ai';
import { z } from 'zod';

export async function POST(req: Request) {
  const { messages }: { messages: UIMessage[] } = await req.json();

  const result = streamText({
    model: 'openai/gpt-4o',
    messages: await convertToModelMessages(messages),
    tools: {
      getWeather: tool({
        description: 'Get the current weather for a location',
        inputSchema: z.object({
          city: z.string().describe('The city name'),
        }),
        execute: async ({ city }) => {
          // Fetch weather data
          const response = await fetch(
            `https://api.weatherapi.com/v1/current.json?key=${process.env.WEATHER_API_KEY}&q=${city}`,
          );
          const data = await response.json();
          
          return {
            temperature: data.current.temp_c,
            condition: data.current.condition.text,
          };
        },
      }),
    },
  });

  return result.toUIMessageStreamResponse();
}

Handle images and files in chat.

'use client';

import { useChat } from '@ai-sdk/react';
import { DefaultChatTransport } from 'ai';
import { useState, useRef } from 'react';
import Image from 'next/image';

export default function Chat() {
  const [files, setFiles] = useState<FileList | undefined>();
  const fileInputRef = useRef<HTMLInputElement>(null);
  
  const { messages, input, handleInputChange, sendMessage } = useChat({
    transport: new DefaultChatTransport({ api: '/api/chat' }),
  });

  const handleSubmit = async (e: React.FormEvent) => {
    e.preventDefault();

    const fileParts = files
      ? await convertFilesToDataURLs(files)
      : [];

    sendMessage({
      role: 'user',
      parts: [{ type: 'text', text: input }, ...fileParts],
    });

    setFiles(undefined);
    if (fileInputRef.current) fileInputRef.current.value = '';
  };

  return (
    <div>
      {messages.map(m => (
        <div key={m.id}>
          {m.role}: 
          {m.parts.map((part, i) => {
            if (part.type === 'text') return <span key={i}>{part.text}</span>;
            if (part.type === 'file' && part.mediaType?.startsWith('image/'))
              return <Image key={i} src={part.url} width={300} height={300} alt="" />;
          })}
        </div>
      ))}

      <form onSubmit={handleSubmit}>
        <input
          type="file"
          accept="image/*"
          onChange={e => setFiles(e.target.files || undefined)}
          ref={fileInputRef}
        />
        <input
          value={input}
          onChange={handleInputChange}
          placeholder="Say something..."
        />
        <button type="submit">Send</button>
      </form>
    </div>
  );
}

async function convertFilesToDataURLs(files: FileList) {
  return Promise.all(
    Array.from(files).map(
      file =>
        new Promise<{ type: 'file'; mediaType: string; url: string }>(
          (resolve, reject) => {
            const reader = new FileReader();
            reader.onload = () =>
              resolve({
                type: 'file',
                mediaType: file.type,
                url: reader.result as string,
              });
            reader.onerror = reject;
            reader.readAsDataURL(file);
          },
        ),
    ),
  );
}

6. Server Actions

Call AI functions directly from Server Components.

import { generateText } from 'ai';

async function generateSummary(text: string) {
  'use server';
  
  const { text: summary } = await generateText({
    model: 'openai/gpt-4o',
    prompt: `Summarize this text: ${text}`,
  });
  
  return summary;
}

export default function Page() {
  return (
    <form
      action={async (formData: FormData) => {
        'use server';
        const text = formData.get('text') as string;
        const summary = await generateSummary(text);
        console.log(summary);
      }}
    >
      <textarea name="text" />
      <button type="submit">Summarize</button>
    </form>
  );
}

Advanced Patterns

Rate Limiting

Implement rate limiting with Upstash:

import { Ratelimit } from '@upstash/ratelimit';
import { Redis } from '@upstash/redis';
import { streamText } from 'ai';

const ratelimit = new Ratelimit({
  redis: Redis.fromEnv(),
  limiter: Ratelimit.slidingWindow(10, '10 s'),
});

export async function POST(req: Request) {
  const ip = req.headers.get('x-forwarded-for') ?? 'anonymous';
  const { success } = await ratelimit.limit(ip);

  if (!success) {
    return new Response('Rate limit exceeded', { status: 429 });
  }

  // Continue with AI logic...
}

Caching Responses

import { streamText } from 'ai';
import { kv } from '@vercel/kv';

export async function POST(req: Request) {
  const { messages } = await req.json();
  const cacheKey = `chat:${JSON.stringify(messages)}`;

  // Check cache
  const cached = await kv.get(cacheKey);
  if (cached) {
    return Response.json(cached);
  }

  const result = await streamText({
    model: 'openai/gpt-4o',
    messages,
  });

  // Cache for 1 hour
  await kv.set(cacheKey, result, { ex: 3600 });

  return result.toUIMessageStreamResponse();
}

Authentication

Protect routes with authentication:

import { auth } from '@/lib/auth';
import { streamText } from 'ai';

export async function POST(req: Request) {
  const session = await auth();
  
  if (!session?.user) {
    return new Response('Unauthorized', { status: 401 });
  }

  // Continue with AI logic...
}

Deployment

Environment Variables

Set in Vercel Dashboard or .env.local:

AI_GATEWAY_API_KEY=your_key
OPENAI_API_KEY=your_key  # if using direct provider access

Deploy to Vercel

pnpm install -g vercel
vercel

Edge Runtime

Optimize for global performance:

import { streamText } from 'ai';

export const runtime = 'edge';

export async function POST(req: Request) {
  // Your AI logic
}

Best Practices

Use maxDuration: Set timeout for API routes
```
export const maxDuration = 30;
```

Handle Errors: Provide user-friendly error messages

try {
  const result = await generateText(...);
} catch (error) {
  return Response.json({ error: 'Failed to generate' }, { status: 500 });
}

Type Safety: Use TypeScript for better DX

interface ChatMessage {
  role: 'user' | 'assistant';
  content: string;
}

Loading States: Show feedback during generation
```
{isLoading ? <Spinner /> : <Result />}
```

Streaming: Use streaming for better UX

const result = streamText({ ... });
return result.toUIMessageStreamResponse();

Guides

Frameworks

Next.js with AI SDK

Next.js with AI SDK

Why Next.js?

Prerequisites

Quick Start

Core Patterns

1. Text Generation

2. Streaming Chat

3. Structured Output

4. Tool Calling

6. Server Actions

Advanced Patterns

Rate Limiting

Caching Responses

Authentication

Deployment

Environment Variables

Deploy to Vercel

Edge Runtime

Best Practices

Examples

Resources

Guides

Frameworks

​Next.js with AI SDK

​Why Next.js?

​Prerequisites

​Quick Start

​Core Patterns

​1. Text Generation

​2. Streaming Chat

​3. Structured Output

​4. Tool Calling

​5. Multi-Modal Input

​6. Server Actions

​Advanced Patterns

​Rate Limiting

​Caching Responses

​Authentication

​Deployment

​Environment Variables

​Deploy to Vercel

​Edge Runtime

​Best Practices

​Examples

​Resources

Next.js with AI SDK

Why Next.js?

Prerequisites

Quick Start

Core Patterns

1. Text Generation

2. Streaming Chat

3. Structured Output

4. Tool Calling

5. Multi-Modal Input

6. Server Actions

Advanced Patterns

Rate Limiting

Caching Responses

Authentication

Deployment

Environment Variables

Deploy to Vercel

Edge Runtime

Best Practices

Examples

Resources