ChatGPT-GUI

ChatGPT Graphical User Interface

A Qt-based graphical user interface application for ChatGPT API. Both ChatGPT-3 and ChatGPT-4 are supported.

Developer: Eliran Wong

Background

We integrated ChatGPT in one of our gui applications, UniqueBible.app. Here in this project, we modify the codes to make ChatGPT-GUI as a standalone application for wider purposes.

Cross-platform

Winodws, macOS, Linux, ChromeOS are supported.

You may also run on Android via Termux.

ChatGPT API

This application is Qt-based graphical user interface that uses OpenAI ChatGPT API to generate chat conversations.

Users need to register an OpenAI account and generate a API key first.

Read pricing at: https://openai.com/pricing

Generate API key at: https://platform.openai.com/account/api-keys

For ChatGPT-4 Users

ChatGPT-GUI supports both ChatGPT-3 and ChatGPT-4.

ChatGPT-GUI uses 'gpt-3-turbo' by default. To use ChatGPT-4:

1. Make sure you have access to 'gpt-4' or 'gpt4-32k' with your OpenAI API key
2. Select 'gpt-4' or 'gpt4-32k' as API Model in 'Chat Settings'
3. Enter BOTH 'OpenAI API Key' AND 'Organization ID' in 'Chat Settings'

Difference between "ChatGPT-GUI" interface and ChatGPT web version

ChatGPT web version is available at: https://chat.openai.com/chat

"ChatGPT-GUI" uses the same model, but with enhanced features not available at ChatGPT web version.

With "ChatGPT-GUI", users can:

• include latest internet search results in ChatGPT responses

• enter multiline-message

• predefine context for conversations. With "ChatGPT-GUI", users can specify a context for conversations. For example, enter "talk about English literature" as the chat context in "Chat Settings", to get ChatGPT responses related to "English literature". In addition, users can can choose to apply their customised contexts only in the beginning of a chat or all inputs.

• adjust temperature [What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.]

• adjust number of choices in ChatGPT responses [How many chat completion choices to generate for each input message.]

• adjust font size for text display

• use python plugins, to automate tasks, add predefined context, or to process ChatGPT responses before they are displayed

• edit, print and save conversations

• save conversations for offline use

• search history, based on title or content

• perform search and replace on chat content

• oranize chat history into different separate database files

• enter message with voice-typing

• use OpenAI image model to generate images

• support system tray

• choose to use regular expression for search and replace

Download and Setup (Windows)

Run in terminal:

git clone https://github.com/eliranwong/ChatGPT-GUI.git

cd ChatGPT-GUI

python -m venv venv

.\venv\Scripts\activate

pip3 install PySide6 openai gtts pyqtdarktheme duckduckgo-search

Install 'pocketsphinx' to support voice-typing [optional]:

pip3 install pocketsphinx

Alternatives to PySide6 [optional]:

"ChatGPT-GUI" supports either PySide6, PySide2 or PyQt5. You may use PySide2 or PyQt5 instead of PySide6, run

pip3 install qtpy PySide2 PyQt5

In config.py, change "qtLibrary" to either "pyside2" or "pyqt5", e.g.:

qtLibrary = 'pyside2'

Run (Windows)

Run in terminal:

cd ChatGPT-GUI

.\venv\Scripts\activate

python ChatGPT-GUI.py

Download and Setup (macOS, Linux, ChromeOS)

Run in terminal:

git clone https://github.com/eliranwong/ChatGPT-GUI.git

cd ChatGPT-GUI

python3 -m venv venv

source venv/bin/activate

pip3 install PySide6 openai gtts pyqtdarktheme duckduckgo-search

Install 'pocketsphinx' to support voice-typing [optional]:

pip3 install pocketsphinx

Alternatives to PySide6 [optional]:

"ChatGPT-GUI" supports either PySide6, PySide2 or PyQt5. You may use PySide2 or PyQt5 instead of PySide6, run

pip3 install qtpy PySide2 PyQt5

In config.py, change "qtLibrary" to either "pyside2" or "pyqt5", e.g.:

qtLibrary = 'pyside2'

Run (macOS, Linux, ChromeOS)

Run in terminal:

cd ChatGPT-GUI

source venv/bin/activate

python3 ChatGPT-GUI.py

QuickStart

1. Launch ChatGPT-GUI by running "python3 ChatGPT-GUI.py"
2. Click the "Settings" button to enter your own OpenAI API key
3. Enter your message and click the "Send" button, to start a conversation

Graphical User Interface

The ChatGPT-GUI Window is separated into left and right areas, by a movable splitter. Users can move the splitter to either end to hide one of them.

On the Left, from top to bottom:

• search entry and button for title - enter text here to search for a title, that contains the entered text, in saved record. You may either press the button or press the "Enter" key after text entry.

• search entry and button for content - enter text here to search for a content, that contains the entered text, in saved record. You may either press the button or press the "Enter" key after text entry.

• record list - list of saved records or search results. Select a record to update the content view on the right.

• "Remove" & "Clear All" buttons - remove selected record or remove all records. Confirmation is required before you proceed the deletion.

• "Help" button - open this wiki page

On the right, from top to bottom:

• message input - users enter a message here to begin or continue a conversation. The voice checkbox toggle voice-typing. The send button sends the message. Alternately, users can press the "Enter" key after text entry. When a message is sent, a progress bar is displayed to indicate that the message is being processed.

• + / - button - toggle multiline user input for message entry

• chat or image - this determines whether the message to sent to generate a chat response or an image.

• content view - display conversation content

• Search and Replace - update content with search and replace; "Repace" button replaces selected text only; "ALL" button replaces all strings that matches search input.

• Control interface:

• Settings - open "Chat Settings" dialog, read more below.

• Temperature - according to OpenAI documentation, temperature means "What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic."

• Choices - according to OpenAI documentation, choices means "How many chat completion choices to generate for each input message."

• Font - edit the font size of chat record list and content view

• Editable checkbox - toggles to edit the content view or make it read-only

• "New" - open a new conversation

• "Save" saves edited conversation. Each conservation is automatically saved when a new response is loaded. Users may edit and updated the content with this button.

Application menu:

• New Database - create a new database file

• Open Database - open an existing database file

• Save Database as - duplicate the currently opened database a different file

• New Chat - create a new chat

• Save Chat - save chat content

• Print Chat - print chat content

• Configure - open settings dialog

• Toggle Dark Theme - toggle between dark and light theme

• Toggle System Tray - toggle system tray; restart the application is required

• Toggle Regular Expression - toggle regular expression for search and replace

• Exit - Exit the application

Chat Settings

• OpenAI API Key - enter your OpenAI key

• Organization ID - enter your organization ID; required for ChatGPT-4

• API Model - select 'gpt-3.5-turbo', 'gpt-4' or 'gpt4-32k'

• Max Token - specify maximum token

• Predefined Context - '[none]' means do not apply any additional context; '[custom]' means apply applying context specified in 'Chat Context' described below; users can add more predefined context via pugins, read more below.

• Chat Context - enter customized context

• Apply Context in - 'First Input Only' or 'All Inputs'

• Latest Online Search Results - check to include; uncheck to exclude

• Maximum Onine Search Results - specify the maximum number of online search results to be included

Include Latest Internet Search Results

OpenAI ChatGPT is currently limited to data up until September 2021. Therefore, ChatGPT itself does not have access to real-time information.

ChatGPT-GUI offers users an option to include the latest internet search results, powered by DuckDuckGo search engine, in ChaptGPT responses.

DuckDuckGo is a search engine that emphasizes user privacy and security, and does not store or track user data like most other search engines do. It was launched in 2008 and has gained popularity among users who are concerned about their online privacy.

You can find the option in "Chat Settings":

When latest internet search results are not included:

When latest internet search results are included:

FAQ - Frequently Asked Questions

How to edit a record's title?

Edit the first line of a conversation record to change its title.

Where are database files stored?

Chat database file is stored in folder "chats" by default. The default file is named as "default.chat". The latest version of ChatGPT-GUI can store database files where users like.

What is the format of the database files?

We use ".chat" as an extension for database files that work with the GUI. They are all standard sqlite files.

Can I change the context to make the conversation to focus on a particular area?

Yes. Click the "Settings" button to edit the chat context. The default is empty. Change it to the topic you want to focus in your conversations.

Can I change the language of the interface?

Yes, simply edit the 'thisTranslation' entry in file 'config.py'.

How to use python plugins to process ChatGPT responses before they are displayed?

It can be done via use of plugins

First, create a python file and save it in folder "plugins".

For example, print response on console in addition to displayint it on GUI:

import config

def printOnConsole(text):
print(text) return text

config.chatGPTTransformers.append(printOnConsole)

You may even modify the content before it is displayed on the GUI.

In the python file, append the method that transform the response to config.chatGPTTransformers

For example, change all characters to upper cases:

import config

def convertToUpperCases(text):
return text.upper()

config.chatGPTTransformers.append(convertToUpperCases)

For example, ChatGPT is weak to produce responses in traditional Chinese. The following plugin ensure that all simplified Chinese characters are converted into traditional Chinese characters:

import config
from opencc import OpenCC

def convertToTraditionalChinese(text):
cc = OpenCC('s2t')
return cc.convert(text)

config.chatGPTTransformers.append(convertToTraditionalChinese)

How to use plugins to customize input suggestion?

You can customize input suggestions by modifying 'config.inputSuggestions' with use of plugins.

For example:

1. Save a python file, e.g. inputSuggestions.py, in folder "plugins".
2. Add content, for example:

import config

config.inputSuggestions = config.inputSuggestions + ["Write a summary", "Write an outline", "Write a letter"]

How to use plugins to customize predefined contexts?

You can customize predefined contexts by modifying 'config.predefinedContexts' with use of plugins.

For example:

1. Save a python file, e.g. predefinedContexts.py, in folder "plugins".
2. Add content, for example:

import config

config.predefinedContexts["Introduction"] = """Write a introduction pertaining to the following content."""

config.predefinedContexts["Summary"] = """Write a summary pertaining to the following content."""

You may even use awesome prompts suggested at https://github.com/f/awesome-chatgpt-prompts

For examples:

import config

config.predefinedContexts["Linux Terminal"] = """I want you to act as a linux terminal. I will type commands and you will reply with what the terminal should show. I want you to only reply with the terminal output inside one unique code block, and nothing else. do not write explanations. do not type commands unless I instruct you to do so. When I need to tell you something in English, I will do so by putting text inside curly brackets {like this}."""

config.predefinedContexts["English Translator and Improver"] = """I want you to act as an English translator, spelling corrector and improver. I will speak to you in any language and you will detect the language, translate it and answer in the corrected and improved version of my text, in English. I want you to replace my simplified A0-level words and sentences with more beautiful and elegant, upper level English words and sentences. Keep the meaning same, but make them more literary. I want you to only reply the correction, the improvements and nothing else, do not write explanations." """

config.predefinedContexts["Spoken English Teacher"] = """I want you to act as a spoken English teacher and improver. I will speak to you in English and you will reply to me in English to practice my spoken English. I want you to keep your reply neat, limiting the reply to 100 words. I want you to strictly correct my grammar mistakes, typos, and factual errors. I want you to ask me a question in your reply. Now let's start practicing, you could ask me a question first. Remember, I want you to strictly correct my grammar mistakes, typos, and factual errors."""

config.predefinedContexts["Biblical Translator"] = """I want you to act as an biblical translator. I will speak to you in english and you will translate it and answer in the corrected and improved version of my text, in a biblical dialect. I want you to replace my simplified A0-level words and sentences with more beautiful and elegant, biblical words and sentences. Keep the meaning same. I want you to only reply the correction, the improvements and nothing else, do not write explanations."""