Xiaozhi Esp32
Xiaozhi Esp32 is an open-source AI hardware chatbot platform for ESP32-based voice interaction, emotional companionship, and smart home control.
Last verified:
What is Xiaozhi Esp32?
Xiaozhi Esp32 is a hand-built AI hardware chatbot platform centered on voice interaction, device management, and configurable character behavior. It is designed to turn supported hardware into a conversational assistant that can be woken by voice, connect to a server, and respond with speech. The website positions it as an open-source AI hardware project that can be used for emotional companionship, smart home control, in-vehicle robots, and other embedded voice products.
The system includes a web console for managing devices, adding new devices with a verification code, and configuring each device's nickname, voice, and personality traits. It also supports a history page where users can review conversations, listen to prior audio, and add speaker profiles so the device can better distinguish between different people. The help pages also show device status through RGB light behavior, which helps users understand whether the device is in pairing, listening, speaking, recording, or updating states.
Xiaozhi Esp32 is aimed at makers, hobbyists, developers, and companies building low-cost AI voice products on supported chipsets. The site and related documentation emphasize ESP32-based hardware and also mention support across multiple chip families, suggesting it is intended for embedded AI devices rather than general desktop use. The platform is especially relevant to users who want a ready-made voice chatbot framework with custom roles, speech interaction, and device provisioning.
The website also frames the service as easy to get started with: connect the device to Wi-Fi, add it in the console, and then configure behavior from the backend. The FAQ and help pages stress privacy limits, wake-word behavior, and the ability to restore default settings or manage speaker profiles. Overall, it is a voice-first AI hardware console plus firmware ecosystem for building and operating connected AI companion devices.
Xiaozhi Esp32 pricing
Pricing model: Freemium
The website says the chatbot is currently free to use. The help page states that it is free at the moment, but future limits or charges may be added depending on compute usage. It also says that if you build your own server or use a third-party API, those costs are separate and not covered by the free service. The site does not show a tiered pricing table on the pages reviewed, so the visible pricing model is a free current service with a possible future paid or restricted model.
Xiaozhi Esp32 pros
- Open-source AI hardware ecosystem
- Voice wake-word interaction
- Device management console
- Add devices with verification code
- Configure nickname and personality
- Select different robot voices
- Speaker recognition support
- Conversation history viewer
- Audio playback for past chats
- Restore default device settings
- RGB status indicators on device
- Works with 2.4G Wi-Fi setup
- 4G version supports plug-and-play use
- Supports OTA firmware updates
- Free current usage on website
- Suitable for makers and embedded projects
- Targets emotional companion use cases
- Useful for smart home scenarios
- Applicable to in-vehicle robot products
- Supports multiple chip families
Xiaozhi Esp32 cons
- Wi-Fi setup supports only 2.4G
- Cannot interrupt the robot while speaking
- Wake-word required before voice chat
- No continuous listening after power on
- No continuous internet connection after power on
- Speaker recognition accuracy drops with many speakers
- Privacy depends on server-side speech processing
- Service may become limited or paid later
Frequently asked questions about Xiaozhi Esp32
How do I connect XiaoZhi to Wi-Fi?
Power on the device, connect your phone or computer to the device hotspot named something like Xiaozhi-XXXXXX, then open http://192.168.4.1 in a browser. On the setup page, choose a 2.4G Wi-Fi network, enter the password, and tap Connect. If the credentials are correct, the device shows Done and restarts automatically after a short delay.
How do I add a device to the console?
After the device is online, wake it with the phrase 'Hello, XiaoZhi' or '你好，小智'. The device will show a 6-digit verification code prompt. Log in to the xiaozhi.me console, open Device Management, choose Add Device, enter that code, and the device will activate and appear in your dashboard.
What can I change in the device configuration?
In the console, you can open Configure Role for a device and adjust the device nickname, voice selection, and personality or role traits. The page also lets you save the configuration or restore the original default values if you want to undo the changes.
What does the speaker feature do?
The speaker feature lets you record voice profiles tied to a nickname and description. You can open History, review a recent conversation, select a voice clip longer than 3 seconds, and add a speaker profile so the device can better tell people apart in later conversations.
Can the robot recognize different people by voice?
Yes, the site says it can use speaker profiles to distinguish different speakers. You add these profiles from conversation history, and the device can then try to match future voices to those profiles. The help page also notes that recognition becomes less accurate as the number of speakers increases.
Can I interrupt the robot while it is talking?
Not yet. The help page says interruption is not supported at the moment and that speaking is only possible when the red light is on. The page says this capability may be added in a future version.
Does XiaoZhi keep listening all the time?
No. The help page says the device stays in a wake-up waiting state after power-on and does not keep listening continuously. It only starts sending audio data after you wake it with the wake word.
Does XiaoZhi stay connected to the internet all the time?
No. The help page says the device does not stay online continuously after power-on. It remains idle until it is awakened, and only then does it connect to the server for the conversation.
Is the service free?
Yes, it is currently free according to the help page. The site also warns that free usage may change later based on compute costs, and that self-hosted servers or third-party APIs are outside the free online service.
How can I tell what the device is doing?
The RGB light on the device indicates its state. Blue flashing means Wi-Fi setup or network connection, green indicates connected or speech playback depending on the state, and red means the device is recording. The help page also mentions blue behavior during OTA firmware updates.