Connect ESP32 With ChatGPT: Building Smarter IoT Systems

Introduction

This article explains why connecting ESP32 with ChatGPT is useful, how the communication works, how to get an API key, troubleshooting common errors, and what applications can be developed using this powerful combination.

The ESP32 microcontroller is a popular choice for IoT projects thanks to its built-in Wi-Fi and Bluetooth capabilities. ChatGPT, developed by OpenAI, is an advanced AI model that understands and generates human-like text. When you connect ESP32 with ChatGPT, you can create IoT systems that respond to natural language, making automation more intuitive and intelligent.

Why Connect ESP32 with ChatGPT?

Traditionally, ESP32 devices are controlled using dashboards, mobile apps, or physical switches. While effective, these methods often require fixed commands or a technical interface. By connecting ESP32 with ChatGPT, users can interact with their devices in plain language.

For example, instead of pressing a button, you could type or say:

“Turn on the pump for 10 minutes.”
“Show me today’s temperature and humidity.”
“Switch off the light when I leave home.”

This natural communication opens the door for applications like smart homes, healthcare reminders, voice-controlled appliances, and AI-driven monitoring systems.

How ESP32 Communicates with ChatGPT

The ESP32 itself cannot run ChatGPT directly, but it can send and receive data via the internet. The general process is:

User Command: A command is given through text or voice input.
API Request: ESP32 sends the input to ChatGPT via OpenAI’s API.
AI Processing: ChatGPT interprets the command and prepares a response.
Action Execution: ESP32 receives the response and acts accordingly, such as controlling a GPIO pin or logging data.

How to Get an OpenAI API Key

Before connecting ESP32 with ChatGPT, you need an API key from OpenAI. Follow these steps:

Go to https://platform.openai.com and sign up with your email, Google, or Microsoft account.
Verify your email and phone number.
Open the API Keys page: https://platform.openai.com/account/api-keys.
Click “Create new secret key”.
Copy the key (it looks like sk-...) and store it safely. You will not be able to view it again later.
Paste this key into your ESP32 code: String apiKey = "sk-your_api_key_here";

Managing Usage

New accounts get free credits, but once they expire, you need to add a payment method.
You can set usage limits to control costs.
Keep your API key private; if exposed, revoke and regenerate it immediately.

Example Setup to Connect ESP32 with ChatGPT

Requirements

ESP32 development board
Wi-Fi connection
Arduino IDE or PlatformIO
OpenAI API key

Basic Steps

Generate API Key as shown above.
Program ESP32: Use Arduino libraries like WiFiClientSecure and HTTPClient to make HTTPS requests.
Send User Input: Pass commands in JSON format to ChatGPT’s API.
Parse Response: Extract the reply and trigger actions on ESP32.

How to Program ESP32 using Arduino IDE:

Connect the ESP32 to your PC via USB.
Install the Arduino IDE if you don’t have. Open the Arduino IDE.
Install ESP32 boards in arduino IDE if not installed.
Select your ESP32 board under Tools → Board and correct Port.
Copy-paste the provided code.
Upload the sketch to your ESP32.

Here’s a simple Arduino IDE example:

//connect esp32 with chatgpt
#include <WiFi.h>
#include <HTTPClient.h>

const char* ssid = "YOUR_WIFI_SSID";
const char* password = "YOUR_WIFI_PASSWORD";
String apiKey = "YOUR_OPENAI_API_KEY";

void setup() {
  Serial.begin(115200);
  WiFi.begin(ssid, password);
  while (WiFi.status() != WL_CONNECTED) {
    delay(1000);
    Serial.println("Connecting...");
  }
  Serial.println("Connected to WiFi");
  
  if(WiFi.status() == WL_CONNECTED) {
    HTTPClient http;
    http.begin("https://api.openai.com/v1/chat/completions");
    http.addHeader("Content-Type", "application/json");
    http.addHeader("Authorization", "Bearer " + apiKey);

    String requestBody = "{\"model\": \"gpt-3.5-turbo\", \"messages\": [{\"role\": \"user\", \"content\": \"Turn on LED\"}]}";

    int httpResponseCode = http.POST(requestBody);
    if(httpResponseCode > 0) {
      String response = http.getString();
      Serial.println(response);
      // Parse response and take action on GPIO
    } else {
      Serial.println("Error sending request");
    }
    http.end();
  }
}

void loop() {}

//connect esp32 with chatgpt
#include <WiFi.h>
#include <HTTPClient.h>

const char* ssid = "YOUR_WIFI_SSID";
const char* password = "YOUR_WIFI_PASSWORD";
String apiKey = "YOUR_OPENAI_API_KEY";

void setup() {
  Serial.begin(115200);
  WiFi.begin(ssid, password);
  while (WiFi.status() != WL_CONNECTED) {
    delay(1000);
    Serial.println("Connecting...");
  }
  Serial.println("Connected to WiFi");
  
  if(WiFi.status() == WL_CONNECTED) {
    HTTPClient http;
    http.begin("https://api.openai.com/v1/chat/completions");
    http.addHeader("Content-Type", "application/json");
    http.addHeader("Authorization", "Bearer " + apiKey);

    String requestBody = "{\"model\": \"gpt-3.5-turbo\", \"messages\": [{\"role\": \"user\", \"content\": \"Turn on LED\"}]}";

    int httpResponseCode = http.POST(requestBody);
    if(httpResponseCode > 0) {
      String response = http.getString();
      Serial.println(response);
      // Parse response and take action on GPIO
    } else {
      Serial.println("Error sending request");
    }
    http.end();
  }
}

void loop() {}

Expected response on Serial Monitor:

{
  "id": "chatcmpl-9ABC123xyz",
  "object": "chat.completion",
  "created": 1725542345,
  "model": "gpt-3.5-turbo-0125",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Sure! Turning on the LED now."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 20,
    "completion_tokens": 12,
    "total_tokens": 32
  },
  "system_fingerprint": "fp_1234567890"
}

{
  "id": "chatcmpl-9ABC123xyz",
  "object": "chat.completion",
  "created": 1725542345,
  "model": "gpt-3.5-turbo-0125",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Sure! Turning on the LED now."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 20,
    "completion_tokens": 12,
    "total_tokens": 32
  },
  "system_fingerprint": "fp_1234567890"
}

Troubleshooting API Errors

When connecting ESP32 with ChatGPT, you may encounter common API errors. Here’s how to solve them:

Insufficient Quota (insufficient_quota)
- This means you have used up your free credits.
- Fix: Go to billing settings, add a payment method, and set a monthly usage limit.
Invalid API Key (invalid_api_key)
- This happens if your key is missing or entered incorrectly.
- Fix: Double-check your code and make sure the key starts with sk-.
Rate Limit Exceeded (rate_limit_exceeded)
- Too many requests were sent in a short time.
- Fix: Add delays or reduce request frequency.
Network Errors
- If ESP32 fails to connect, check your Wi-Fi connection and SSL certificates.
- Fix: Make sure Wi-Fi credentials are correct and your board has stable internet.

Applications of Connecting ESP32 with ChatGPT

Smart Home Assistants: Control lights, fans, and appliances via natural language.
IoT Monitoring Systems: Get human-like summaries of sensor data.
Healthcare Devices: Send medication reminders and respond to patient queries.
Learning Tools: Build educational kits that respond conversationally to students.

Conclusion

When you connect ESP32 with ChatGPT, you unlock the potential of conversational IoT. This integration makes devices smarter and user-friendly, enabling real-world applications in automation, education, and healthcare.

As AI and IoT continue to converge, the combination of ESP32 and ChatGPT will play a major role in shaping the next generation of interactive systems.

LEARN HOW TO GET STARTED WITH FREERTOS ON ESP32:

Introduction To FreeRTOS On ESP32 – ArduinoYard

ESP32 FreeRTOS Task Priorities And Scheduling Explained: A Basic Guide – ArduinoYard

Connect ESP32 with ChatGPT: Building Smarter IoT Systems