chatgpt mirai qq bot Download - chatgpt mirai qq bot Source code download

chatgpt mirai qq bot

Other source code

v2.5.3 - 维护性更新

Download

cover

ChatGPT for Bot

A robot that supports chatting in various mainstream language models!

» View the tutorial »

Discord Group 1, QQ Group 2, QQ Group 3, QQ Group 4, QQ Group 5, and QQ Development Group will release the latest project updates, video tutorials, Q&A and QQ questions and exchanges. Before joining the group, please see if the content here can solve your problem.
If it cannot be solved, prepare the problems, logs and configuration files before asking questions.
There are many ChatGPT QQ robots in the debugging group, which do not answer technical questions.

⚡ Support

? Multi-platform compatibility

We support multiple chat platforms.

platform	Group chat reply	Private chat reply	Conditional trigger	Administrator commands	Drawing	Voice reply
Mirai	support	support	support	support	support	support
OneBot	support	support	support	support	support	support
Telegram	support	support	Partial support	Partial support	support	support
Discord	support	support	Partial support	Not supported	support	support
Enterprise WeChat	support	support	support	Not supported	support	support
Personal WeChat	support	support	support	Not supported	support	support

? Order

You can learn about the internal commands of the robot in the Wiki.

? Build

If you are a mobile phone party, you can read this mobile phone deployment tutorial (using a Linux server): https://www.bilibili.com/video/av949514538

AidLux: Deploy using only old Android phones

Run the following command to start the automatic installation script.

bash -c " $( wget -O- https://gist.githubusercontent.com/B17w153/f77c2726c4eca4e05b488f9af58823a5/raw/4410356eba091d3259c48506fb68112e68db729b/install_bot_aidlux.sh ) "

Deployment Tutorial

Linux: Deployment through rapid deployment scripts (recommended by newcomers)

Execute the following command to start the automatic deployment script. It installs Docker, Docker Compose and writes configuration files for you.

bash -c " $( wget -O- https://gist.githubusercontent.com/lss233/2fdd75be3f0724739368d0dcd9d1367d/raw/62a790da4a391af096074b3355c2c2b7ecab3c28/chatgpt-mirai-installer-gocqhttp.sh ) "

Linux: Deployment via Docker Compose (with Mirai)

We use `docker-compose.yaml` to integrate [lss233/mirai-http](https://github.com/lss233/mirai-http-docker) and this project for rapid deployment. But there are still some steps to configure during the deployment process.

You can view the construction tutorial on the Wiki.

Linux: Deployment via Docker (suitable for users who already have Mirai)

Find a suitable location and write your config.cfg .
Execute the following command to start the bot:

 # 修改 /path/to/config.cfg 为你 config.cfg 的位置
# XPRA_PASSWORD=123456 中的 123456 是你的 Xpra 密码，建议修改
docker run --name mirai-chatgpt-bot 
    -v /path/to/config.cfg:/app/config.cfg 
    --network host 
    lss233/chatgpt-mirai-qq-bot:browser-version

Windows: Quick deployment package (with Mirai/go-cqhttp, recommended by newcomers)

We have created a quick boot package for Windows users, which can be found in Release.

The file name is: quickstart-windows-go-cqhttp-amd64.zip (recommended) or quickstart-windows-mirai-amd64.zip

Mac: Quick deployment package (with its own Mirai, recommended by newcomers)

Mac users can also use the Windows Quick Deployment Package. @magisk317 has been tested and the functions are basically normal. However, it should be noted that if you need to use the image mode, since wkhtmltoimage.exe cannot run on Mac, you can use wkhtmltopdf instead and install the command:

 brew install --cask wkhtmltopdf

Please refer to the link for details on the installation and usage of brew.

Manual deployment

Tip: You need Python >= 3.11 to run this project

Deploy Mirai and install the mirai-http-api plugin.
Download this project:

git clone https://github.com/lss233/chatgpt-mirai-qq-bot
cd chatgpt-mirai-qq-bot
pip3 install -r requirements.txt

Adjust the configuration file according to the project documentation.
Start bot.

python3 bot.py

[Advertising] Free OpenAI API Key

You can get free OpenAI API Key tests here.

? HTTP API

Additional HTTP API support will be provided after adding the following configuration in `config.cfg`.

[ http ]
# 填写提供服务的端口
host = " 0.0.0.0 "
port = 8080
debug = false

After startup, the following interface will be provided:

POST /v1/chat

Request parameters

Parameter name	Must-choose	type	illustrate
session_id	yes	String	Session ID, default: `friend-default_session`
username	yes	String	Username, default:`某人`
message	yes	String	Message cannot be empty

Request Example

{
    "session_id" : " friend-123456 " ,
    "username" : " testuser " ,
    "message" : " ping "
}

Response format

Parameter name	type	illustrate
result	String	SUCESS, DONE, FAILED
message	String[]	Text return, support multi-part return
voice	String[]	Audio returns, supports base64 encoding of multiple audios; reference: data: audio/mpeg; base64,,iVBORw0KGgoAAAANS...
image	String[]	Image return, supports base64 encoding of multiple images; reference: data:image/png; base64, UhEUgAAAgAAAIACAIA...

Response Example

{
    "result" : " DONE " ,
    "message" : [ " pong! " ],
    "voice" : [],
    "image" : []
}

POST /v2/chat

Request parameters

Parameter name	Must-choose	type	illustrate
session_id	yes	String	Session ID, default: `friend-default_session`
username	yes	String	Username, default:`某人`
message	yes	String	Message cannot be empty

Request Example

{
    "session_id" : " friend-123456 " ,
    "username" : " testuser " ,
    "message" : " ping "
}

Please note that session_id should be in the canonical format. Its format is friend- or group- plus string

Example

 friend-R6sxRvblulTZqNC
group-M3jpvxv26mKVM

If it cannot be correct, it will be handled according to the group

Response format String: request_id

Response Example

 1681525479905

Note that the returned content may be quoted. Please remove the quotes. (Including " and ' )

 '1681525479905'

GET /v2/chat/response

Request parameters

Parameter name	Must-choose	type	illustrate
request_id	yes	String	Request id, value returned by /v2/chat

Request Example

 /v2/chat/response?request_id=1681525479905

Please note that request_id cannot be quoted (including " and ' ). The following is an example of error

 /v2/chat/response?request_id='1681525479905'

 /v2/chat/response?request_id="1681525479905"

 /v2/chat/response?request_id='1681525479905"

 /v2/chat/response?request_id="1681525479905'

Response format

Parameter name	type	illustrate
result	String	SUCESS, DONE, FAILED
message	String[]	Text return, support multi-part return
voice	String[]	Audio returns, supports base64 encoding of multiple audios; reference: data: audio/mpeg; base64,,iVBORw0KGgoAAAANS...
image	String[]	Image return, supports base64 encoding of multiple images; reference: data:image/png; base64, UhEUgAAAgAAAIACAIA...

Each request returns increments and clears. There are no more returns after DONE and FAILED.

Response Example

{
    "result" : " DONE " ,
    "message" : [ " pong! " ],
    "voice" : [ " data:audio/mpeg;base64,iVBORw0KGgoAAAANS... " ],
    "image" : [ " data:image/png;base64,UhEUgAAAgAAAAIACAIA... " , " data:image/png;base64,UhEUgAAAgAAAAIACAIA... " ]
}

Please note that when returning SUCCESS , it means waiting

{ "result" : " SUCCESS " , "message" : [], "voice" : [], "image" : []}

Note that there may be multiple DONE , please keep requesting until FAILED appears. FAILED indicates that the reply is completed.

{ "result" : " FAILED " , "message" : [ " u6ca1u6709u66f4u591au4e86uff01 " ], "voice" : [], "image" : []}

Please note that SUCCESS may be interspersed between DONE and FAILED . The entire reply cycle may be greater than one minute.

? Loading presets

If you want the robot to bring a certain chat style automatically, you can use the preset function.

We have two presets:猫娘and正常. You can learn how to write presets in the presets folder.

Use加载预设猫娘to load the cat girl preset.

Here are some preset short videos, you can see the effect:

MOSS: https://www.bilibili.com/video/av309604568
Ding Zhen: https://www.bilibili.com/video/av267013053
Xiao Heizi: https://www.bilibili.com/video/av309604568
Gao Qiqiang: https://www.bilibili.com/video/av779555493

Detailed tutorial on preset systems: Wiki

You can get presets shared by everyone at Awesome ChatGPT QQ Presets.

You can also refer to Awesome-ChatGPT-prompts-ZH_CN to train your ChatGPT, and you can also refer to Awesome ChatGPT Prompts to unlock more skills.

? Text to pictures

When sending code or sending a message to a QQ group fails, the message is automatically converted to a picture and sent.

Font files are stored in fonts/ directory.

The default font is bolder.

? Text to pronunciation

Since v2.2.5, we support access to Microsoft's Azure engine and VITS engine to let your robot send voice.

Tip : To use voice functions on Windows platforms, you need to install the latest VC runtime library, which you can download here. `

? Similar projects

If you have the idea of making robots yourself, you can take a look at the following projects:

Ariadne - An elegant and complete Python QQ robot framework (mostly this!!!)
mirai-api-http - Provides HTTP API for all languages to use mirai QQ bot
Reverse Engineered ChatGPT by OpenAI - Unofficial ChatGPT Python Support Library

This project is developed based on the above projects, so you can also click a star for them!

Besides us, there are some great projects: