Connect ChatGPT to iMessage: Complete Guide (2026)

Want ChatGPT to respond to your iMessages? With Sendblue and the OpenAI API, you can build a ChatGPT-powered iMessage agent in under an hour. Messages come in as blue bubbles, ChatGPT processes them, and responses go back as iMessages.

How It Works

The setup uses three components:

Sendblue — Handles iMessage sending/receiving via API and webhooks
OpenAI API — Processes messages with GPT-4 or GPT-4o
Your server — A simple Express app that connects the two

When someone texts your Sendblue number, the message hits your webhook. Your server sends it to ChatGPT, gets a response, and sends it back via Sendblue. The entire round trip takes 2-5 seconds.

Setup

npm install express sendblue openai

You need API keys from Sendblue (free) and OpenAI.

The Code

import express from 'express';
import Sendblue from 'sendblue';
import OpenAI from 'openai';

const app = express();
app.use(express.json());

const sendblue = new Sendblue(
  process.env.SENDBLUE_API_KEY,
  process.env.SENDBLUE_API_SECRET
);
const openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY });

const threads = new Map(); // conversation history per number

app.post('/webhook', async (req, res) => {
  const { from_number, content } = req.body;

  if (!threads.has(from_number)) {
    threads.set(from_number, [
      { role: 'system', content: 'You are a helpful assistant responding via iMessage. Keep responses brief and conversational.' }
    ]);
  }

  const messages = threads.get(from_number);
  messages.push({ role: 'user', content });

  // Show typing indicator
  await sendblue.sendTypingIndicator({ number: from_number });

  const completion = await openai.chat.completions.create({
    model: 'gpt-4o',
    messages,
    max_tokens: 500,
  });

  const reply = completion.choices[0].message.content;
  messages.push({ role: 'assistant', content: reply });

  await sendblue.sendMessage({
    number: from_number,
    content: reply,
  });

  res.json({ ok: true });
});

app.listen(3000);

Customization Ideas

This basic setup can be extended in many ways:

Business-specific system prompts: Tell ChatGPT about your products, pricing, and policies so it can answer customer questions accurately
Image processing: When users send photos via iMessage, forward the media_url from the webhook to GPT-4 Vision for analysis
Tool calling: Use OpenAI function calling to let ChatGPT book appointments, check order status, or look up information in your database
Scheduled messages: Have ChatGPT send daily summaries, reminders, or check-ins at scheduled times

For a more advanced AI agent architecture using Claude, see our Claude + Sendblue tutorial.

Why iMessage for ChatGPT?

Other ChatGPT-to-iMessage solutions (like LoopMessage) are limited to personal use. Sendblue's approach gives you:

A real business phone number — not your personal iMessage
Any AI model — not locked to ChatGPT; swap in Claude, Gemini, or local models
Full API control — send media, vCards, typing indicators, read receipts
Scale — handle thousands of conversations, not just your own texts
Compliance — SOC 2, HIPAA support for business use cases

Get started free — no credit card required.

ChatGPT

iMessage API

Tutorial

Ready to send your first iMessage?

Get API access in minutes. Free sandbox, no credit card required.

Get API Access