You should use the latest version, as it fixes bugs and frequently updates code for improved logic.
Extension will be updated later.

Website

Add-Ons List:
[ChatBot] [Report]

Description

This plugin serves as a library for AI to fetch responses from servers.

It is intended to be used only by AddOns, DLCs, and other plugins.

If the plugin version contains SNAPSHOT, it means that version is a test build and may contain bugs or have reduced performance.

Currently, there is no public hosting platform for the plugin's API library, so developers may need to wait for proper hosting support.

If you have any complaints, please open an issue.

If you have ideas or suggestions, feel free to leave a comment. I’ll implement them if possible.

Command Usage
( Note: These commands do not require permissions since they don't affect other players.)

Set token to database (token is encrypted before saving):

Additional info for set token:
The API provides a method getPlayerToken which returns the player’s token in its encrypted form. It is decrypted only when sending requests to the AI server.

Important:
If you attempt to manually decrypt a player’s token, you may face legal consequences. I, the owner and developer, will not be responsible for any misuse or attempts to reverse-engineer player tokens.

Other Commands

Get list of AddOns:
Get list of DLCs:
Show available platforms:
Show all models from each platform:
Show all models from a specific platform:
config.yml

Code (Text):

enable: false

# Database support
# - sqlite
# - mysql
# - postgresql
# - mongodb
database:
  type: sqlite

  sqlite:
    path: artificialintelligence.db

  mysql:
    host: localhost
    port: "3306"
    name: mcengine
    user: root
    password: ""

  postgresql:
    host: localhost
    port: "5432"
    name: mcengine
    user: postgres
    password: ""

  mongodb:
    uri: mongodb://localhost:27017
    name: mcengine
    collection: artificialintelligence

token:
  # Used to encrypt player tokens before saving them to the database
  # Must have a length of 16 characters
  secretKey:

ai:
  custom:
    # You can define multiple servers under "custom"
    # Each server can have its own URL, token, and default model

    server1:
      # API endpoint URL for server1
      url: http://localhost:11434/v1/chat/completions
      # Authorization token for server1
      token:
      # Default model name for server1
      models:
        model1: model1

    server2:
      # API endpoint URL for server2
      url: http://localhost:11435/v1/chat/completions
      # Authorization token for server2
      token:
      # Default model name for server2
      models:
        model1: model1
        model2: model2

  deepseek:
    token:
    # Default model loaded when the plugin starts
    models:
      model1:

  openai:
    token:
    # Default model loaded when the plugin starts
    models:
      model1:

  openrouter:
    token:
    # Default model loaded when the plugin starts
    models:
      model1:

Important Notes

The API includes two response-fetching methods:
- One uses the server token
- The other uses the player's token

Each plugin, addon, or DLC may use a different method — make sure to check which one is being used.

For multiple model usage:
- If the same model and platform are used, the system will reuse the same instance to avoid duplication and save memory.
- However, if you use multiple models (e.g., `{platform}:{model}`), each will create a separate instance.
This can lead to higher RAM usage. If your server has sufficient RAM, this should not be an issue.

Developer
You can create your own AddOns and DLCs too.
Click here for more information: AddOn DLC
API