Stream GPT-6 Astra responses from a Next.js route handler

What you get

A Next.js App Router endpoint that streams GPT-6 Astra output to the browser token by token using the Responses API.

Prerequisites

An OpenAI API key in OPENAI_API_KEY (e.g. in .env.local)
A Next.js app using the App Router
The openai npm package

Steps

Install the SDK.

Shell

npm install openai

Create app/api/chat/route.ts. It calls client.responses.create with stream: true and forwards only response.output_text.delta events.

TypeScript

import OpenAI from "openai";

const client = new OpenAI(); // reads OPENAI_API_KEY

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

  const stream = await client.responses.create({
    model: "gpt-6-astra",
    input: [{ role: "user", content: prompt }],
    stream: true,
  });

  const encoder = new TextEncoder();
  const body = new ReadableStream({
    async start(controller) {
      for await (const event of stream) {
        if (event.type === "response.output_text.delta") {
          controller.enqueue(encoder.encode(event.delta));
        }
      }
      controller.close();
    },
  });

  return new Response(body, {
    headers: { "Content-Type": "text/plain; charset=utf-8" },
  });
}

Read the stream in a client component.

TypeScript

const res = await fetch("/api/chat", {
  method: "POST",
  body: JSON.stringify({ prompt: "Say 'double bubble bath' ten times fast." }),
});
const reader = res.body!.getReader();
const decoder = new TextDecoder();
while (true) {
  const { done, value } = await reader.read();
  if (done) break;
  setText((t) => t + decoder.decode(value)); // your React state setter
}

Test it

Shell

curl -N -X POST http://localhost:3000/api/chat \
  -H "Content-Type: application/json" \
  -d '{"prompt":"Write a haiku about streaming."}'

Text should appear in chunks rather than all at once.