# 78/xiaozhi-esp32

An MCP-based chatbot | 一个基于MCP的聊天机器人

Repository: https://github.com/78/xiaozhi-esp32
Canonical: https://ross.abutalabs.com/products/xiaozhi-esp32
Homepage: https://xiaozhi.me
Language: C++
License: MIT
License Family: permissive
Topics: chatbot, esp32, mcp
Last push: 2026-08-21T05:52:29+00:00

## Health v2 (maintenance only)
Score: 88/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 96, longevity 52
- inputs: {"age_days": 732, "days_push": 12, "days_rel": 28, "gap_med": 7.0, "n_releases_24m": 59}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 29186, forks 6746 (observed 2026-08-28T04:11:53.146409+00:00)

## What it is
XiaoZhi is an open-source MCP-based AI voice chatbot firmware for ESP32-family microcontrollers, connecting large language models like Qwen and DeepSeek to real-world hardware. It supports offline voice wake-up, streaming ASR/LLM/TTS pipelines, camera vision input, and device/cloud-side MCP control of peripherals such as LEDs, servos, and smart home devices.

## Use cases
- build a diy ai voice assistant on esp32
- control smart home devices with a voice chatbot
- give an llm camera vision on embedded hardware
- create a custom wake word voice assistant
- connect mcp ai agents to physical hardware
- build an ai companion device with oled display

## When to choose
- you want an open-source voice assistant running fully on ESP32 hardware
- you need MCP-based device control bridging LLMs and real-world peripherals
- you want offline wake-word detection with streaming ASR/LLM/TTS
- you are a hobbyist following the extensive DIY board tutorials

## When to avoid
- you need a pure software chatbot without embedded hardware
- your project targets non-ESP32 microcontrollers
- you want a polished consumer product rather than a DIY firmware project
- you need cloud-independent operation without any LLM API access

## Facets
- artifact type: application
- maturity: active
- function: chatbot, speech-recognition, tts, llm-inference, mcp, agent-framework, audio-processing, computer-vision, iot, embedded
- domain: artificial-intelligence, chatbots, speech-processing, iot, embedded-systems, hardware
- platform: embedded, iot, cpp
- tags: esp32, mcp-protocol, voice-assistant, voice-wake-word, opus-audio, websocket, mqtt, smart-home, hardware-diy, oled-display, ai-agents

## Member repositories
- 78/xiaozhi-esp32 (main) score 88

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:53.146409+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:53:09.092091+00:00, confidence not recorded.
  - readme: https://github.com/78/xiaozhi-esp32 (fetched 2026-08-28T04:11:53.146409+00:00, sha 83936ba384d2)
  - homepage: https://xiaozhi.me (fetched 2026-08-29T07:50:16.193828+00:00, sha e1b43f11f135)
- Data as of 2026-08-30T08:39:29.467469+00:00.