PSC.CSharp.Library.ChatGPT 8.0.4

ChatGPT library for C#

This is another C# library for ChatGPT using official OpenAI API that allows developers to access ChatGPT, a chat-based large language model. With this API, developers can send queries to ChatGPT and receive responses in real-time, making it easy to integrate ChatGPT into their own applications.

For more details about this project, see my following posts:

• New ChatGPT library for C#
• Write a ChatGPT client
• GitHub repository

Quick example

using PSC.CSharp.Library.ChatGPT;

var chat = new ChatGpt("<API_KEY>");
var response = await chat.Ask("What is the weather like today?");
Console.WriteLine(response);

Table of Contents

• Features
• Getting Started
  • Usage
    • ChatGPT Official API
• Configuration options
  • ChatGPT Official API
• Examples
  • ChatGPT Console App
  • Use a different model
  • Using ChatGPT Official API For Free

Features

• Easy to use.
• Using official OpenAI API.
• Supports both free and pro accounts.
• Supports multiple accounts, and multiple conversations.
• Support response streaming, so you can get response while the model is still generating it.

Getting Started

To install PSC.CSharp.Library.ChatGPT, run the following command in the Package Manager Console:

Install-Package PSC.CSharp.Library.ChatGPT

Alternatively, you can install it using the .NET Core command-line interface:

dotnet add package PSC.CSharp.Library.ChatGPT

Usage

ChatGPT Official API

Here is a sample code showing how to use PSC.CSharp.Library.ChatGPT:

using PSC.CSharp.Library.ChatGPT;

// ChatGPT Official API
var chat = new ChatGpt("<API_KEY>");

// get response
var response = await chat.Ask("What is the weather like today?");
Console.WriteLine(response);

// stream response
await chat.AskStream(response => {
    Console.WriteLine(response);
}, "What is the weather like today?");

// get response for a specific conversation
var response = await chat.Ask("What is the weather like today?", "conversation name");
Console.WriteLine(response);

// stream response for a specific conversation
await chat.AskStream(response => {
    Console.WriteLine(response);
}, "What is the weather like today?", "conversation name");

Configuration options

ChatGPT Official API

ChatGptOptions
{
    string    BaseUrl;           // Default: https://api.openai.com
    double    FrequencyPenalty;  // Default: 0.0;
    long      MaxTokens;         // Default: 64;
    string    Model;             // Default: gpt-3.5-turbo
    double    PresencePenalty;   // Default: 0.0;
    string[]? Stop;              // Default: null;
    double    Temperature;       // Default: 0.9;
    double    TopP;              // Default: 1.0;
}

BaseUrl

By default, the base URL is https://api.openai.com. You can set the BaseUrl to a free reverse proxy server to use ChatGPT Official API for free.

Frequency Penalty

How much to penalize new tokens based on their existing frequency in the text so far. Decreases the model's likelihood to repeat the same line verbatim.

Max Tokens

The maximum number of tokens to <strong>generate</strong> is shared between the prompt and completion. The exact limit varies by model. (One token is roughly 4 characters for standard English text)

Model

The model to use. The default is gpt-3.5-turbo. The ModelValue class contains the popular models to use.

• gpt-3.5-turbo
• gpt-3.5-turbo-16k
• gpt-4-turbo

You can also set the Model with any other model name. For more details, see the OpenAI API Model documentation.

Presence Penalty

How much to penalize new tokens based on whether they appear in the text so far. Increases the model's likelihood to talk about new topics.

Stop

Up to four sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

Temperature

Controls randomness: Lowering results in less random completions. As the temperature approaches zero, the model will become deterministic and repetitive.

TopP

Controls diversity via nucleus sampling: 0.5 means half of all likelihood-weighted options are considered.

Examples

ChatGPT Console App

This is a simple console app that uses PSC.CSharp.Library.ChatGPT to interact with ChatGPT.

using PSC.CSharp.Library.ChatGPT;

// ChatGPT Official API
var chat = new ChatGpt("<API_KEY>");

var prompt = string.Empty;

while (true)
{
    Console.Write("You: ");
    prompt = Console.ReadLine();

    if (prompt is null) break;
    if (string.IsNullOrWhiteSpace(prompt)) break;
    if (prompt == "exit") break;

    Console.Write("ChatGPT: ");
    await chat.AskStream(Console.Write, prompt, "default");
    Console.WriteLine();
}

Use a different model

You can use a different model by passing the model name to the constructor.

var chat = new ChatGpt("<API_KEY>", new ChatGptOptions
{
    Model = "gpt-3.5-turbo"
});

Using ChatGPT Official API For Free

you can use ChatGPT Official API by setting the base URL to a free reverse proxy server.

var chat = new ChatGpt("<API_KEY>", new ChatGptOptions
{
    BaseUrl = "https://api.youreverseproxy.com"
});

Errors

When a call to ChatGPT fails, there are several possible errors that can be returned. The full error will be in the Error property. A common example is

{
  "choices": null,
  "created": 0,
  "error": {
    "message": "You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors.",
    "type": "insufficient_quota",
    "param": null,
    "code": "insufficient_quota"
  },
  "id": null,
  "model": null,
  "object": null,
  "usage": null,
  "success": false
}

insufficient_quota

This is a very common error. That means you haven't add a payment details on your account, you haven't add money in the account or you have reached the limit of your quote.