Skip to main content

Function: GetAssistantModelByProvider()

GetAssistantModelByProvider(options): typeof VercelAi | typeof OpenAIAssistant | typeof DeepSeekAssistant | typeof GoogleAIAssistant | typeof XaiAssistant | typeof OllamaAssistant | typeof AnthropicAssistant

Defined in: packages/core/src/lib/model-utils.ts:40

Returns the appropriate Assistant model based on the provider. (Internal use)

Parameters​

options​

The options object

chatEndpoint?​

string

The chat endpoint that handles the chat requests, e.g. '/api/chat'. This is required for server-side support.

provider?​

string

The name of the AI provider. The supported providers are: 'openai', 'anthropic', 'google', 'deepseek', 'xai', 'ollama'

Returns​

typeof VercelAi | typeof OpenAIAssistant | typeof DeepSeekAssistant | typeof GoogleAIAssistant | typeof XaiAssistant | typeof OllamaAssistant | typeof AnthropicAssistant

The assistant model class.

Example​

import { GetAssistantModelByProvider } from '@openassistant/core';

const AssistantModel = GetAssistantModelByProvider({
  provider: 'openai',
});

// configure the assistant model
AssistantModel.configure({
  apiKey: 'your-api-key',
  model: 'gpt-4o',
});

// initialize the assistant model
const assistant = await AssistantModel.getInstance();

// send a message to the assistant
const result = await assistant.processTextMessage({
  text: 'Hello, world!',
});