@venturialstd/openai

A lightweight Package for OpenAI, developed by Venturial, that allows you to interact with OpenAI's Responses API. Designed to be used by directly instantiating a client with dynamic configuration, without requiring a global module. Ideal for projects under the Venturial organization.

Features

• Generate responses using OpenAI's Responses API.
• Persist OpenAI responses in a TypeORM-managed database.
• Dynamically provide API key and model settings via Venturial's SettingsService.
• Fully compatible with NestJS HttpModule, SharedModule, and provider injection.
• Preconfigured with default models like gpt-3.5-turbo, gpt-4, and gpt-4-turbo.

Installation

npm install @venturialstd/openai
# or
yarn add @venturialstd/openai

Basic Usage

1. Use the OpenAI client in your service

import { Injectable } from "@nestjs/common";
import { OpenaiResponseService } from "@venturialstd/openai";

@Injectable()
export class HarveyAIService {
  constructor(private readonly openaiService: OpenaiResponseService) {}

  async askQuestion(input: string, instructions: string, model: string) {
    const response = await this.openaiService.createResponse(
      input,
      instructions,
      model
    );
    return response.output_text;
  }
}

API

OpenAIClient | OpenaiResponseService

| Method | Parameters | Description | | ---------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------- | ------------------------------------------------------------------------- | | getClient() | none | Returns a configured OpenAI client and default model from settings. | | createResponse(input: string \| Responses.ResponseInput, instructions: string, model: string, store?: boolean) | input, instructions, model, optional store (default true) | Creates a new OpenAI response and optionally persists it in the database. |

Settings

The module supports dynamic configuration via Venturial's SettingsService. Default settings:

| Key | Type | Description | Default | | ------------------------------- | ------- | ------------------------------- | --------------- | | GLOBAL:OPENAI:GENERAL:ENABLED | Boolean | Enable or disable OpenAI module | false | | GLOBAL:OPENAI:GENERAL:API_KEY | Secret | OpenAI API Key | "" | | GLOBAL:OPENAI:GENERAL:MODEL | Select | Default OpenAI Model | "gpt-3.5-turbo" |

Available models:

• gpt-3.5-turbo
• gpt-4
• gpt-4-32k
• gpt-4-turbo
• gpt-4-turbo-16k
• gpt-4-turbo-32k
• gpt-4-turbo-8k
• gpt-4.1-nano

Notes

• Requires openai, @nestjs/common, @nestjs/axios, and @venturialstd/core as peer dependencies.
• Works with both dynamic database settings and hardcoded API keys.
• Perfect for projects where AI responses must be stored or audited.
• Developed and maintained by Venturial for their internal and client projects.

License

MIT