ESP32-S31 Development Platform
ESP32-S31 provides dual-core RISC-V processing at up to 320 MHz and integrates Wi-Fi 6, Bluetooth 5.4 (LE + BR/EDR), Thread 1.4, and Zigbee 3.0, making it a hardware foundation for networked embedded projects.
ESP-Mosaico
ESP-Mosaico 是一款基于 ESP32-S31 的模块化 AIoT 与嵌入式开发板，结合设备端 Agent、触控界面和可扩展外设。它面向希望通过自然语言或 PC 端 AI 编程 Agent 构建、测试和部署实体设备应用的创作者与开发者。
A Development Board for AI-Assisted Embedded Creation
ESP-Mosaico is a modular development device for AIoT and embedded creation, based on Espressif ESP32-S31. It brings an on-device Agent, touch interface, network connectivity, and plug-in peripherals together on one hardware platform, allowing developers to iterate from idea validation to operation on a physical device.
The product supports two related workflows: Chat Coding uses natural language to create or adjust functions on the device, while Vibe Coding uses an AI coding Agent on a PC together with a simulator, debugging tools, and the physical device for coding, deployment, testing, and feedback. Expansion interfaces can connect modules such as an RGB matrix light for projects with physical interaction and visual effects.
Key Features
Touch Interface and Expansion Interfaces
The device features a 480 × 480 OLED touch display, a Type-C high-speed USB 2.0 port, and two 20-pin expansion interfaces. The factory-installed ESP-Claw firmware integrates touch applications, onboard peripherals, and the on-device Agent runtime.
On-Device Conversational Programming
In Chat Coding, users can send text requests through a bound WeChat account or another supported platform, or enter voice input on the device. ESP-Claw provides built-in skill types including quick questions, function creation, skill search, memory, planning, and scheduled tasks.
AI Development Workflow for Physical Devices
Vibe Coding combines an AI coding Agent on a PC with repository documentation, the ESP-GSP simulator, debugging tools, and the physical device to create a development loop. UI and interactions can be verified first and then deployed to the physical device for testing.
Modular Peripheral Expansion
Expansion modules can connect peripherals. The 8×8 RGB matrix light module contains 64 WS2812C-2020-V6 RGB pixels for pixel patterns, animations, and status feedback. Modules can be inserted on the left or right side, and the mainboard identifies the module and installation side through EEPROM.
Use Cases
Create Device Functions with Natural Language
Describe a small function or automation requirement through Chat Coding, then save the function completed by the Agent to Work for repeated execution or further modification. Suitable for creators who want to quickly experiment with screen graphics, scheduled tasks, or Lua automation on the device.
Develop and Debug Embedded Applications
Provide an AI coding Agent on a PC with the esp-mosaico-vibe repository and a specific task, such as implementing Snake. First verify the UI, interactions, and visual effects in the ESP-GSP simulator, then flash the result to the physical device for adjustment.
Connect Custom Hardware and I/O
Use the two 20-pin expansion interfaces to connect peripherals. Follow the interface definitions for power, USB-Serial-JTAG, UART, or pins that can be initialized as general-purpose GPIO to build sensor, actuator, or other I/O projects.
Create Pixel Lighting Effects and Status Displays
Insert the 8×8 RGB matrix light module into the side of the mainboard to create pixel animations, expressions, icons, simple information displays, or operating-status prompts. The accompanying program can control 64 RGB pixels, and the mainboard can identify the installation side.
Build Audio Visualization Prototypes
Use the Bluetooth audio application in the factory-installed firmware to connect to a phone for audio playback, and combine it with the optional 8×8 light matrix to run a music visualization application. Microphone mode can also convert ambient sound into real-time lighting effects.
Pros and Cons
Pros
- Brings an on-device Agent, touch applications, onboard peripherals, and expandable hardware together on one development platform.
- Supports both conversational creation and a PC-based AI Agent development workflow, with the latter allowing UI and interaction verification in a simulator first.
- Provides wireless connectivity through Wi-Fi 6, Bluetooth 5.4, Thread 1.4, and Zigbee 3.0.
- The factory-installed firmware includes touch applications such as camera, gallery, settings, and Bluetooth audio, and supports skills, Lua extensions, and MCP capabilities.
Cons
- The Chat Coding text or voice workflow depends on Wi-Fi, a bound supported platform, and available ASR and LLM services; online AI features also require the corresponding service configuration.
- The 8×8 RGB matrix light module has a display resolution of 8 × 8 per module, making it better suited to simple patterns, status prompts, and pixel-style content.
- The camera expansion module can only use the left slot, and the interface orientation must be checked when installing an expansion board.
FAQ
How do I get started with Vibe Coding?
You can use an AI coding Agent on a PC together with the esp-mosaico-vibe repository, ESP-GSP simulator, debugging tools, and the physical device to complete development, deployment, and testing. The guide lists Codex, Claude Code, OpenCode, and Trae as examples of optional Agents; connect the device to the computer using a USB cable that supports data transfer.
What needs to be prepared before using Chat Coding?
First, install or open Mosaico-Claw Idea in Vibe Mode, complete Wi-Fi configuration, and bind the device to WeChat or another supported platform. Send text conversations through the bound platform, or press and hold the input field in the device's Claw app to enter voice input. Chat Coding also requires available ASR and LLM services.
How are AI and network services configured?
On the same local network, open the control panel using the IP address shown on the device settings page to configure Wi-Fi, LLM, ASR, network API services, and some skill and Lua options. The device must be restarted manually after changing settings. The ASR and LLM keys provided from the factory support the basic experience; configure your own model credentials when the quota is used up or stronger coding performance is needed.
How do I install the 8 × 8 RGB matrix light module?
The RGB matrix light module can be plugged into the left or right side of the mainboard. Its EEPROM allows the mainboard to identify the module type and installation side. Power off the device before installation, insert the module into the lower row of the mainboard's dual-row side connector, and confirm the orientation. Low brightness is recommended for initial testing.
Quick Facts
- Category
- AIoT and embedded development board
- Core platform
- Espressif ESP32-S31
- Processor
- Dual-core RISC-V, up to 320 MHz
- Display and interfaces
- 480 × 480 OLED touch display; two 20-pin expansion interfaces; Type-C high-speed USB 2.0
- Wireless connectivity
- Wi-Fi 6, Bluetooth 5.4 (LE + BR/EDR), Thread 1.4, Zigbee 3.0
- Development workflows
- Chat Coding and Vibe Coding
Alternatives à ESP-Mosaico
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
GitBot
GitBot is a local browser-based workspace for creating reusable coding bots powered by Claude Code, Codex, or OpenCode. It helps developers run those agents across repository folders with persistent threads and configurable permissions.
hob
hob is an independent workspace for coding agents that keeps agent sessions, terminals, history, and follow-up work organized around the tools and providers you already use. It is aimed at developers who want local control over routing, history, and workspace structure rather than a bundled model stack.
Manta AI
Manta AI is an autonomous web app testing tool for teams that want to map application behavior, catch regressions, and generate tests without writing scripts or maintaining selectors. It works from a URL and supports plain-English test flows, run results with screenshots, and scheduled or deployment-triggered checks.
SonOf
SonOf connects to your repo and PM tool, audits the codebase and surrounding product context, and turns approved work into shipped tickets with senior engineering review. It is aimed at founders and engineering leaders who need backlog help without hiring a full team immediately.