> For the complete documentation index, see [llms.txt](https://rocket-networking.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://rocket-networking.gitbook.io/docs/server-side-scripting/ai-functions-on-the-server.md).

# AI Functions on the Server

The `callAI` method is a server-side function available within the `Game` class. This method allows you to make requests to the Rocket Networking AI service, powered by OpenAI, to generate AI-based responses using the GPT-3.5 Turbo model. You can call this function in every `Game` instance to interact with the AI service.

### Method Signature

```javascript
async callAI(messagesArray: Array<{ role: string, content: string }>, temperature: number): Promise<any>
```

### Parameters

* **messagesArray** (Array of objects): An array of message objects that form the conversation context. Each message object should have a `role` (either 'system', 'user', or 'assistant') and `content` containing the text of the message.
* **temperature** (number): A floating-point value controlling the randomness of the AI's response. Higher values (e.g., 0.8) make the response more random, while lower values (e.g., 0.2) make it more deterministic.

### Returns

* A Promise that resolves to the AI response data.

### Usage

You can use the `callAI` method within a `Game` instance to make AI requests as follows:

```javascript

const messages = [
  { role: "system", content: "You are a helpful assistant." },
  { role: "user", content: "Tell me a joke." },
  { role: "assistant", content: "Why did the chicken cross the road?" }
];
const temperature = 0.7;

game.callAI(messages, temperature)
  .then((response) => {
    console.log("AI response data is:");
    console.log(response);

    if(response == "error"){
      //error
    }else{
      // Handle the AI response here. response.message is the reply and response.creditsUsed is how many credits were used
    }
  })

```

### Error Handling

* If there's an error in making the AI request (e.g., network issues), the method will catch the error and return the string "error."