ESP32 + AI
Voice assistants, AI gadgets and on-device machine learning.

Ivy AI: Fully Offline Voice Assistant on ESP32-S3
Build a standalone voice assistant on a $20 ESP32-S3 board. Wake word, speech-to-text, on-device LLM, and TTS — no cloud, no server.

Build an ESP32 Alexa Smart Speaker with I2S Audio
Turn an ESP32 into a working Alexa voice client using an I2S mic, I2S codec, and Neopixel status LEDs. Full wiring, pin table, and build steps.

Dsn AI Voice Assistant v2: ESP32-C3 Build Guide
Build a pocket-sized AI voice assistant with ESP32-C3, INMP441 mic, MAX98357A amp, and Hugging Face cloud AI. Streams WAV audio in real time.
Forrest Note: AI Voice Notes on ESP32-S3 with Whisper & GPT
Build a pocket voice-note device on an ESP32-S3 e-ink board. Records speech, transcribes with Whisper, rewrites with GPT-4o-mini, and syncs Markdown to GitHub and Obsidian.

ESP32 AI Monitor: Desk Display for Claude, ChatGPT and More
Turn a Cheap Yellow Display into a live AI usage dashboard for Claude, ChatGPT, Gemini, Copilot and Cursor — no Wi-Fi needed, just USB serial.

Orion: Build a Voice Assistant on ESP32-S3 You Own
Flash an ESP32-S3 board in one double-click and get a private desk voice assistant that speaks Arabic and English with your own API keys.

esphome-devices: 12 Ready-to-Adopt ESPHome Configs for ESP32
Voice assistants, e-paper calendars, LVGL touch dimmers and HUB75 matrices — real running ESPHome configs with honest notes on what works.

ESP32-S3-WATCH-rs: Build a Voice-Driven Smartwatch OS in Rust
Flash a bare-metal Rust OS onto a Waveshare ESP32-S3 smartwatch. Get wake-word detection, smart home control, media playback and 50+ API endpoints.

ESP32-S3 AI Voice Assistant with Gemini Live API
Build a programmable voice interface on ESP32-S3 with wake-word detection, function calling, and Gemini Live API integration for real conversational AI.

boardsmith: Text Prompt to KiCad Schematic and ESP32 Firmware
Turn a plain-English description into a wired KiCad schematic, BOM, and ESP32 firmware in seconds — no templates, real computed values.

Nest Audio PCB: Open-Source ESP32-S3 Voice Assistant Drop-In
Replace your Google Nest Audio mainboard with this open-source ESP32-S3 PCB for a fully local Home Assistant voice assistant.

ESP32-S3 AI Voice Assistant with Gemini, STT and TTS
Build a full AI voice assistant on ESP32-S3 using Google Cloud Speech, Gemini AI and Text-to-Speech — 63% faster with dual-core parallel processing.

esp32-tinyLLM: Run a 28.9M-Parameter LLM on ESP32-S3
Flash a Rust inference engine for a 28.9M-parameter language model onto an ESP32-S3. Generates children's stories at 10.5 tok/s — faster than the C port it beat.

ESPHome Voice Satellite on Waveshare ESP32-S3-AUDIO-Board
Turn the Waveshare ESP32-S3-AUDIO-Board into a Home Assistant voice satellite with on-device wake word, LED ring, and music mixing — all in pure ESPHome.

ESP32 Voice Assistant with Gemini, Whisper and Piper TTS
Build a talking AI assistant on ESP32 using Faster-Whisper for speech recognition, Gemini 2.5 Flash for responses, and Piper TTS for natural voice output.

Rotary Phone Voice Assistant with ESP32 and Home Assistant
Turn a vintage Siemens W48 rotary phone into an OpenAI-powered voice assistant using ESP32 and ESPHome. Lift the handset, speak, get answers.

NucleoOS: Web-Native OS with Offline AI for M5Stack Cardputer
Flash NucleoOS onto the M5Stack Cardputer for a full multi-app OS with offline AI, bilingual voice, a browser desktop shell, games, and a security lab.

ESP32MCPServer: Connect AI Agents to Real-World Sensors
Build an MCP server on ESP32 that lets AI assistants query live I2C sensors, NMEA GPS, NMEA 2000 marine bus, and OBD-II car diagnostics over WebSocket.

ESP-Assistant: ESP32-S3 Voice Assistant for Home Assistant
Build a compact ESP32-S3 voice assistant (VoicePuck) that runs on-device wake word detection and integrates natively with Home Assistant via ESPHome.

Build a Free AI Voice Assistant with ESP32-S3 and HuggingFace
Build a fully free AI voice assistant on an ESP32-S3 board. Speech-to-text, LLM, and TTS all run on a free HuggingFace Space — no paid APIs needed.

Oído: Open-Vocabulary Speech Recognition on a $5 ESP32-S3
Run a quantized Conformer-CTC recognizer entirely on an ESP32-S3 — 3.7% LibriSpeech WER, no cloud, no NPU, English and Spanish supported.

Tokpet: ESP32-S3 Desk Pet That Watches Your AI Token Usage
Build a round AMOLED ESP32-S3 desk pet that shows your Claude, Codex, and DeepSeek usage as mood-driven rings. Local-first, zero-config LAN discovery.

Koala Satellite: DIY ESP32-S3 Voice Assistant for Home Assistant
Build a far-field voice satellite for Home Assistant using ESP32-S3 DevkitC and ReSpeaker Lite with XMOS XU316 noise cancellation, LED strip, and rotary encoder.

Embedded-AI-Harness: AI Closed-Loop Firmware Dev for ESP32
Let Claude write, flash, and test ESP32 firmware on real hardware in a closed loop — spec to working silicon, hands off. Built on a Raspberry Pi testbench.

DIY Video Doorbell with Voice Response (ESP32 Cam + ESPHome)
Build a cloudless video doorbell with live camera, actionable phone notifications, and pre-recorded voice replies using ESP32-CAM, DFPlayer Mini, and Home Assistant.

zclaw: Personal AI Assistant on ESP32 in Under 888 KiB
Build a tiny AI assistant on ESP32 with GPIO control, Telegram chat, scheduled tasks, and custom tools — all packed into 888 KiB firmware.