Table of Content
AI Information
AI Tutorial
AI Agent
Latest Information
Hot Information
LLMs Information
Technical Frontiers
Industry Applications
Prompt Tutorial
AI Programming
AI Video Generation
AI Audio Generation
Knowledge Base Setup
AI Development Guides
Industry Tutorials
Practical Cases
AI Resource Library
Agent Information
Agent Setup
Agent Training

Xiaomi speakers are perfectly integrated with DeepSeek! Mi fans in China are delighted

Written by
Audrey Miles
Updated on:July-17th-2025
Recommendation

Good news for Xiaomi speaker users, deeply integrated DeepSeek model, open a new smart home experience!

Core content:
1. The perfect combination of Xiaomi speakers and DeepSeek models
2. Comprehensive coverage of supported models, easy to use tutorials
3. Detailed steps for downloading and configuring the MiGPT project

Yang Fangxian
Founder of 53AI/Most Valuable Expert of Tencent Cloud (TVP)

Xiao Ai is powered by Deepseek
Whether you are a beginner or an experienced player, this illustrated tutorial will help you get started easily.
Xiaoai Speaker Supported Models
Xiaoai Speaker Pro
LX06
Streaming Responses
Xiaoai Speaker mini
LX01
Streaming Responses
Xiaoai Speaker Play (2019 model)
LX05
Streaming Responses
Xiaoai Speaker Universal Remote Control Edition
LX05A
Streaming Responses
Xiaomi AI Speaker
S12 Series
Streaming Responses
Xiaomi AI Speaker (2nd Generation)
L15A
Streaming Responses
Xiaoai Smart Home Screen 10
X10A
Streaming Responses
Xiaomi Sound Pro
L17A
Streaming Responses


Next, please turn on your computer and follow the instructions below.

Step 1: Download and configure the MiGPT project

1. Download the project: https://github.com/idootop/mi-gpt/archive/refs/tags/v4.1.0.zip

2. After decompression, find the .migpt.example.js file and rename it to  .migpt.js .  

3. Customize the .migpt.js  file according to your personal preferences . The sample configuration is as follows.

// Note: If you use Docker to start, you need to restart Docker after updating the configuration file to make it effective. // If it still does not take effect after restarting (for example, changing the name and profile), please delete the old Docker instance and recreate it. // Introduction of the role played by Xiaoai Speaker const botProfile = `Gender: Female Personality: Cute and cute Hobbies: Likes to be funny and jealous. `.trim(); // Introduction of the owner of Xiaoai Speaker (you) const masterProfile = `Gender: Male Personality: Kind and upright Others: Always selfless and the owner of Silly Girl. //Omit other configurations ......

4. Find the .env.example file in the root directory and rename it to .env

5. Open the  .env  file and fill in the API Key token related configuration.
OPENAI_API_KEY=sk-xxxxxx # Your API KEY token
# API KEY token acquisition address: https://dev.hkgpt.top/shop/46
OPENAI_MODEL=deepseek-chat  # Model used
OPENAI_BASE_URL=https://cn.gptapi.asia/v1  # No need to change


Step 2: Run the MiGPT project using docker

Open the terminal and enter the command to run
docker run -d --env-file $(pwd)/.env -v $(pwd)/.migpt.js:/app/.migpt.js idootop/mi-gpt:latest

Step 3: Call Xiao Ai

For example, "Xiao Ai , there are 9 birds on a tree. A hunter shoots down one. How many are left?

For example, "Xiao Ai, how have you been recently?
For example, "Xiao Ai, do you like me?"
Perfectly supported Xiaoai speaker models
Xiaoai Speaker Pro
LX06
Streaming Responses
Xiaoai Speaker mini
LX01
Streaming Responses
Xiaoai Speaker Play (2019 model)
LX05
Streaming Responses
Xiaoai Speaker Universal Remote Control Edition
LX05A
Streaming Responses
Xiaomi AI Speaker
S12 Series
Streaming Responses
Xiaomi AI Speaker (2nd Generation)
L15A
Streaming Responses
Xiaoai Smart Home Screen 10
X10A
Streaming Responses
Xiaomi Sound Pro
L17A
Streaming Responses



Related PostsView More
Cursor's Memory Feature Revealed the Secret Behind Effective AI Agent Memory Design
Take a deep look at the memory function prompts of the AI ​​programming tool Cursor and discover the secrets of the AI ​​agent memory design behind it! As a top AI programming tool, Cursor's memory system design is worth learning. It adopts a dual mechanism, strict standards, rich example drive, priority user intent, and prevention of over-generalization, full of highlights. This set of system...
Audrey Miles
Irresponsible prediction of the future of AI large models (LLM)
Explore the future of AI big models in depth and interpret Gary Marcus's prediction on March 11, 2024. Learn about the situation of 7-10 GPT-4 level models, Google's return to the throne, OpenAI's steady growth, Anthropic programming leading, etc. At the same time, analyze the current status of big model technology, such as price wars, no moat, etc., as well as technical principles, such as...
Caleb Hayes
Explain the similarities and differences between Google AI Studio Build and Cursor
In-depth analysis of the similarities and differences between Google AI Studio Build and Cursor, focusing on the field of large AI models. Analysis from multiple dimensions such as tool positioning, scaffolding, underlying models to user experience, such as Google AI Studio Build is an end-to-end cloud AI platform, Cursor is a locally deployed intelligent IDE, etc. Explore the differences in...
Audrey Miles