How to Integrate an AI API in Node.js and Express (Streaming Proxy)

Not every team runs on Next.js. Plenty of startups still ship Express APIs behind a React or Vue SPA, mobile clients, or webhook workers. The integration pattern is the same as other stacks: keep the API key on the server, forward JSON to your AI provider, and stream tokens back to the client.

This tutorial targets Node 20+ with Express. For App Router examples, see Next.js AI API integration.

Setup

npm install express dotenv

.env:

DAYMORA_API_KEY=your_key_here
PORT=3001

Never commit .env. Rotate keys using the workflow in Managing API keys securely.

Basic JSON route

import express from "express";
import "dotenv/config";

const app = express();
app.use(express.json({ limit: "1mb" }));

app.post("/api/chat", async (req, res) => {
  const upstream = await fetch("https://daymora.com/api/v1/chat", {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      API_SECRET: process.env.DAYMORA_API_KEY,
    },
    body: JSON.stringify({
      model: req.body.model ?? "gpt-5",
      messages: req.body.messages,
      stream: false,
    }),
  });

  const text = await upstream.text();
  res.status(upstream.status).send(text);
});

app.listen(process.env.PORT ?? 3001);

Streaming route

Pipe the upstream body through without buffering the full completion:

app.post("/api/chat/stream", async (req, res) => {
  const upstream = await fetch("https://daymora.com/api/v1/chat", {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      API_SECRET: process.env.DAYMORA_API_KEY,
    },
    body: JSON.stringify({
      model: req.body.model ?? "gpt-5",
      messages: req.body.messages,
      stream: true,
    }),
  });

  if (!upstream.ok) {
    res.status(upstream.status).end(await upstream.text());
    return;
  }

  res.setHeader("Content-Type", "text/event-stream");
  res.setHeader("Cache-Control", "no-cache");
  res.setHeader("Connection", "keep-alive");

  const reader = upstream.body.getReader();
  const pump = async () => {
    while (true) {
      const { done, value } = await reader.read();
      if (done) break;
      res.write(value);
    }
    res.end();
  };
  pump().catch(() => res.end());
});

Understand transport tradeoffs in SSE vs WebSockets for AI streaming.

CORS for SPAs

If your React app runs on localhost:5173 and API on localhost:3001, enable CORS only for known origins — not * with credentials. Rate-limit by user id or session to prevent abuse.

Deploy notes

• Use platform secrets for DAYMORA_API_KEY
• Set request timeouts above your p95 model latency
• Log status and duration_ms, not raw prompts in production

Cost planning

Express services often power high-volume webhooks (support tickets, email ingest). Token bills scale with every message. Estimate with How much does an AI chatbot cost? or consider flat-rate if AI is always on (pricing comparison).

Summary

Express is a thin, proven proxy layer for AI APIs: JSON for simple clients, streamed responses for chat UIs, keys in environment variables only. Pair it with a frontend streaming pattern from our React chatbot guide and you have a stack familiar to most JavaScript teams.