How can we implement AI integration using Vercel AI SDK in Next.js 16 Edge Functions?

You can integrate AI in Next.js 16 Edge Functions using the Vercel AI SDK to easily call language models, handle streaming, and serve AI-powered responses at the edge, improving latency and scaling.

To implement AI integration with Vercel AI SDK in Next.js 16 Edge Functions, first install the SDK packages. Then create an Edge API Route or Server Component that imports the SDK functions (like streamText or generateText) and calls an AI model (e.g., Google Gemini). Use the streaming response utilities to send AI-generated results progressively. Set the route’s runtime to edge for low latency worldwide. This leverages Vercel’s AI Gateway to manage provider integration, caching, and scaling transparently.

Code

// app/api/ai/route.ts
import { streamText, StreamingTextResponse } from 'ai';
import { google } from '@ai-sdk/google';

export const runtime = 'edge';

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

  const response = await streamText({
    model: google('models/gemini-1.5-pro-latest'),
    prompt,
  });
  return new StreamingTextResponse(response);
}