Ai Voice Assistant (DeepSeek) With Quick Response
by STEAM-DIY in Circuits > Assistive Tech
67 Views, 2 Favorites, 0 Comments
Ai Voice Assistant (DeepSeek) With Quick Response
.jpg)
🔊🤖 Build Your Own AI Voice Assistant Using ESP32-S3 (N16R8) 🚀
This guide will walk you through creating an AI voice assistant using the ESP32-S3-DevKitC-1 (WROOM N16R8) with speech input/output and visual feedback via an OLED display. Perfect for STEM classrooms and hobbyists!
Supplies
.png)

.png)
.png)
.png)
Required Hardware
1. Core Components
- ESP32-S3 DevKitC-1 (WROOM N16R8)
- INMP441 Digital Microphone
- MAX98357A Digital Amplifier
- Speaker (8Ω 2
3W or 4Ω 23W) - 128x32 I2C OLED Display
- Breadboard (400-hole x2)
- Jumper Wires / Dupont Wires
- USB Type-C Cable
2. Optional Modules (I am not using this)
- ML307R Cat.1 4G Module (for outdoor use without Wi-Fi)
- 6x6mm Tactile Push Buttons (Optional: Volume control, pause/resume)
🔧 Wiring Overview




💻 Firmware Installation




Recommended Firmware:
Use: v1.4.6_bread-compact-wifi.bin (Stable version for OLED 128x32 SSD1306)
- Download from xiaozhi GitHub Releases
- Flash the .bin file using ESP32 Download Tool or esptool.py
- Press RESET after upload to start
🌐 First Boot
- Voice prompt: "WiFi connected mode"
- OLED Output: Shows AI assistant is listening
- Connect your phone to AP: Xiaozhi-XXXX
- Visit in browser: http://192.168.4.1
OLED showing listening prompt)
www.xiaozhi.me
✅ Tips & Troubleshooting
- White screen on OLED? Check SDA/SCL pin connections and 3.3V power
- No response? Ensure microphone GND and L/R are connected properly
- ESP32 not detected? Try a different USB cable or reinstall drivers
🎉 Done!
Video tutrial . For any question
You now have a fully functioning AI Voice Assistant powered by ESP32-S3 and Xiaozhi AI. Use it in your STEM class, demo it at maker fairs, please Subscibe my channel
if you guys have any problem let me know. thanks