Server

Echo's Next.js SDK will handle all authentication logic out of the box.

Configure Echo with the Echo constructor.

import Echo from '@merit-systems/echo-next-sdk';

export const {
  // Echo Auth Routes
  handlers,

  // Server-side utils
  getUser,
  isSignedIn,

  // AI Providers
  openai,
  anthropic,
  google,
} = Echo({
  appId: 'ECHO_APP_ID',
});

By re-exporting the Echo constructor results, you can use throughout your server logic.

The Echo constructor takes the following configuration params.

Handlers

import handlers from '@/echo';

export { GET, POST } = handlers;

Exporting these routes will expose default authentication routes which can be used from the client. In most cases your code will not need to touch these routes.

• api/echo/signin
• api/echo/callback
• api/echo/refresh

Very important - When productionizing your application, you will need to append /api/echo/callback to your app's homepage route when inputting the correct authorized callback Url.

Server Utilities

import { getUser, isSignedIn } from "@/echo";

export default async function Page() {
    const signedIn = await isSignedIn();

    if (!signedIn) {
        return <SignIn />;
    } else {
        const user = await getUser();
        return <Dashboard user={user} />;
    }
}

AI Providers

Echo's Next.js SDK provides a thin wrapper around the Vercel AI SDK. Echo follows their recommended pattern with a small diff, to route through Echo instead of the model provider.

import { openai } from '@ai-sdk/openai';
import { openai } from '@/echo';
import { streamText, UIMessage, convertToModelMessages } from 'ai';

// Allow streaming responses up to 30 seconds
export const maxDuration = 30;

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

  const result = streamText({
    model: openai('gpt-5'),
    messages: convertToModelMessages(messages),
  });

  return result.toUIMessageStreamResponse();
}