【ESP32】【LLM API】Accessing the Xunfei Spark LLM API with ESP32

Accessing the Xunfei Spark LLM API with ESP32

Real-time conversation with Xunfei Spark large language model via ESP32 microcontroller

Introduction

With the rapid development of artificial intelligence technology, large language models (LLMs) are being applied in increasingly diverse scenarios. Combining these powerful cloud-based AI capabilities with embedded devices can bring entirely new interactive experiences to smart hardware projects. This article will detail how to use an ESP32 microcontroller to access the Xunfei Spark model API via WiFi, enabling real-time conversations.

Project Overview

This project implements the following core functions:

1. WiFi Connection Management: Automatically connects to a specified WiFi network
2. User Interface: Enables user question input and AI response display via serial port
3. HTTP Request Encapsulation: Handles HTTP communication with the Xunfei Spark API
4. JSON Data Processing: Builds request bodies and parses response data
5. Streaming Response Handling: Optimizes the reception and display of large responses

This solution is particularly suitable for IoT devices, smart home controllers, educational equipment, and other embedded projects requiring AI conversation capabilities.

Code Structure Breakdown

1. Initialization and Configuration

// WiFi configuration
const char* ssid = "601B";
const char* password = "12345678";

// API configuration
const String apiUrl = "https://spark-api-open.xf-yun.com/v1/chat/completions";
const String apiPassword = "AoqxxxxPspsrWk";
const String appId = "6cxxxf";
String userId = "exxxp";

// Memory optimization configuration
const size_t MAX_RESPONSE_SIZE = 150000;  // Maximum response length
const size_t JSON_BUFFER_SIZE = 225000;   // JSON buffer size
const size_t LINE_LENGTH = 200;           // Characters displayed per line

This section contains network connection parameters, API access credentials, and memory optimization settings. Since ESP32 has limited memory, properly configuring buffer sizes is crucial.

2. User Input Processing

void handleSerialInput() {
   
  static String inputBuffer;
  
  while (Serial.available()) {
   
    char c = Serial.read();
    
    // Display input prompt
    if (!promptDisplayed && inputBuffer.length() == 0) {
   
      Serial.print("\nYour question: ");
      promptDisplayed = true;
    }

    if (c == '\n') {
    // Enter key pressed
      if (!isProcessing && inputBuffer