Espressif ESP32-S3-BOX-3

Overview

The ESP32-S3-BOX-3 is Espressif's flagship AIoT development kit, combining edge AI inference, voice interaction, and touchscreen HMI in a self-contained enclosure. The ESP32-S3 SoC provides dual Xtensa LX7 cores at 240 MHz with hardware AI acceleration, backed by 16 MB Quad flash and 16 MB Octal PSRAM — sufficient for running TensorFlow Lite models, large LVGL GUIs, and offline speech recognition simultaneously.

The enclosure integrates a 2.4-inch 320×240 SPI touchscreen, two digital MEMS microphones for far-field voice detection, and a built-in speaker. The ESP-SR speech recognition framework supports 200+ customizable voice commands, far-field wake-word detection, and interruption-based voice control. ChatGPT cloud integration is also supported via the provided firmware examples.

Expansion is handled by a high-density PCIe-style connector for official and third-party add-on boards (camera, sensors, etc.). The ESP32-S3-BOX-3 supports Wi-Fi, Bluetooth 5 LE, and Matter-over-Wi-Fi/Thread. The complete hardware design (schematics, PCB, 3D shell STEP files) and firmware are open-source on GitHub. It ships as a standalone module — a separate BOX-3B variant includes the DOCK stand, RGB LED module, USB-C cable, and Dupont wires as a bundle.