Skip to main content

​
Interface: CompletionRequest

Defined in: src/lib/models/prompt.ts:131 Represents a message in a completion request to an AI model. Defines the structure for input messages sent to AI models, supporting different roles (user, system, tool, function) and content types including text and images for vision-enabled models. CompletionRequest

​
Examples

// Simple user message
const userMessage: CompletionRequest = {
  role: "user",
  content: "What's the weather like today?"
};

// System message with instructions
const systemMessage: CompletionRequest = {
  role: "system",
  content: "You are a helpful weather assistant. Provide accurate forecasts."
};

// Multimodal message with text and image
const visionMessage: CompletionRequest = {
  role: "user",
  content: [
    { type: "text", text: "What do you see in this image?" },
    {
      type: "image_url",
      image_url: {
        url: "data:image/jpeg;base64,/9j/4AAQSkZJRgABAQEA...", // Base64 encoded image data
        detail: "high"
      }
    }
  ]
};

// Tool response message
const toolResponse: CompletionRequest = {
  role: "tool",
  content: '{"temperature": 72, "condition": "sunny"}',
  tool_call_id: "call_123"
};

​
Properties

​
content

content: string | CompletionRequestContent[]
Defined in: src/lib/models/prompt.ts:133 Message content as text or multimodal array

​
role

role: "function" | "user" | "system" | "tool"
Defined in: src/lib/models/prompt.ts:132 The role of the message sender

​
tool_call_id?

optional tool_call_id: string
Defined in: src/lib/models/prompt.ts:134 ID of the tool call this message responds to (for tool/function roles)