Click the badge below to access the web applications:
Welcome to the unofficial platform for custom web applications powered by OpenAI APIs! This repository hosts a collection of web applications deployed on Streamlit Cloud. These applications leverage the capabilities of OpenAI's powerful language models to provide unique functionalities. Currently, the platform includes three web applications:
The first web application, Talk to GPT, is an interactive chatbot application that allows users to communicate with the model using text messages or speech input. Developed using OpenAI's Chat Completions API, and powered by OpenAI's GPT models (incl. gpt-4-turbo, gpt-4o, and gpt-3.5-turbo) for generating high-quality responses, Whisper model for speech-to-text conversion, and TTS model for text-to-speech audio output, this application offers a conversational experience similar to interacting with a human expert. Users can select between the GPT-3.5 Turbo, GPT-4 Turbo and GPT-4o models based on their needs, and have the option to play the bot's responses in audio format, enhancing the conversational experience with a greater sense of immersion and realism. Additionally, the application provides a variety of built-in prompts that assign roles or personas to the chatbot, providing an effective starting point for each type of conversation, and ensuring that the chatbot will produce the desired responses in an efficient and appropriate manner.
The second web application, CodeMaxGPT, is designed to provide coding assistance to programmers. Built on OpenAI's Assistants API, it is specifically tuned and optimized to cater to the diverse needs of developers, including code generation, debugging, refactoring, and documentation. The platform allows users to choose between the GPT-3.5 Turbo, GPT-4 Turbo and GPT-4o models, providing auto-prompts for the GPT model based on various use cases that a user may select. It also features a more user-friendly interface compared to the original ChatGPT, allowing users to comfortably enter or paste code and even upload code scripts directly from their local computer. With its user-friendly interface, advanced auto-prompting features, and seamless code uploading capabilities, CodeMaxGPT is the ultimate coding companion. Whether you're a junior developer tackling complex programs or a seasoned pro exploring new programming languages or frameworks, CodeMaxGPT has you covered.
Talk to GPT provides the following features:
-
Constructed using with OpenAI's Chat Completions API, enables the selection of either the gpt-3.5-turbo, gpt-4o or gpt-4-turbo model to generate high-quality human-like responses to user’s prompts.
-
Enables communication with the GPT model through either text messages or speech input.
-
Utilizes OpenAI's Whisper model for accurate speech-to-text conversion for user’s speech input, and TTS (text-to-speech) model for chatbot's audio response output.
-
Provides an option to play the bot’s responses in audio format for an immersive and realistic conversational experience.
-
Offers a variety of built-in prompts that assign roles or personas to the chatbot, provides an effective starting point for each type of conversations, and ensures that the chatbot will produce the desired responses in an efficient manner.
Below are the features offered by CodeMaxGPT:
-
Enables the selection of either the gpt-3.5-turbo, gpt-4o or gpt-4-turbo model to generate high-quality human-like responses to user’s prompts.
-
Built upon OpenAI's Assistants API, specifically tuned and optimized to provide coding assistance to programmers of all levels.
-
Supports various coding tasks, including code generation, debugging, refactoring, adding comments, code reviewing, generating GitHub README, and suggesting solutions to coding challenges.
-
Provides auto-prompts for the GPT model tailored to the various coding tasks that users may need help with.
-
Offers a user-friendly interface that surpasses the original ChatGPT.
-
Allows comfortable code entry and pasting, and direct uploading of code scripts from the user’s local computer, for enhanced convenience.
-
Enables users to code confidently in unfamiliar territory.
The repository structure of the project is as follows:
OpenAI-API-Web-Apps/
├── .streamlit/
│ └── config.toml
├── assets/
│ ├── ChatGPT-Tkinter-Desktop-App.exe
│ └── cover-page.gif
├── pages/
│ ├── 2_Talk_To_GPT.py
│ └── 3_CodeMaxGPT.py
├── Home.py
├── packages.txt
├── requirements.txt
├── .gitignore
├── README.md
└── LICENSE
The description of each file and folder in the repository is as follows:
- .streamlit/: This folder contains the config.toml file, which configures the appearance of the Streamlit web application. The config.toml file specifies the theme settings such as primary color, background color, text color, and font.
- assets/: This folder contains additional assets used in the project, including the cover-page.gif image file for the cover page. It also includes the ChatGPT-Tkinter-Desktop-App.exe, which is a simplified desktop version of Talk to GPT. You can find the source code for the desktop application in the ChatGPT-Tkinter-Desktop-App repository.
- pages/: This folder contains the Python code that powers the three web applications. It includes the following Python scripts:
- 2_Talk_To_GPT.py: Python script for the Talk to GPT web application.
- 3_CodeMaxGPT.py: Python script for the CodeMaxGPT web application.
- Home.py: This is a Python script for the home page of the Streamlit web applications. It contains code related to the navigation between the three web applications.
- packages.txt: The file manages the project dependencies and is necessary for deploying the web applications on Streamlit Cloud.
- requirements.txt: This file lists all the required Python modules and packages. It is also necessary for the deployment of the web applications on Streamlit Cloud. It ensures that the required dependencies are installed when deploying the applications.
- .gitignore: This file intentionally specifies untracked files that Git should ignore.
- README.md: Provides an overview of this repository.
- LICENSE: The license file for the project.
The web applications are currently hosted on Streamlit Cloud. To access the applications, please visit OpenAI API Web Applications on Streamlit. Once you visit the URL, you will be presented with a home page where you can select the desired web application. Click on the application you want to use, and it will open in a new tab or window.
To use the Talk to GPT application, follow these steps:
- Visit the Talk to GPT on Streamlit.
- Select your desired GPT model.
- Input your OpenAI API key in the field at the top.
- You can start interacting with the chatbot using either the "MESSAGE BOT" or "TALK TO BOT" options.
- For text input, click on the "MESSAGE BOT" expander, select a built-in prompt from the dropdown menu, and press CTRL + Enter to submit. The chatbot will respond with a generated message. You can continue the conversation by entering your own messages.
- For speech input, make sure the text message input field is cleared, then click on the "TALK TO BOT" expander, click on the microphone icon, and speak your message. The speech input will be converted to text, and the chatbot will respond accordingly.
Below are two GIF images that demonstrate the usage of the Talk to GPT application:
Interacting through text messages
Interacting through speech input
To use the CodeMaxGPT application, follow these steps:
- Visit the Introducing CodeMaxGPT webpage.
- Click on the "Get started now" button, and you'll be directed to the CodeMaxGPT on Streamlit.
- Select your desired GPT model (gpt-4-turbo is recommended).
- Input your OpenAI API key in the field at the top.
- You can now start interacting with the coding assistant by entering your text message or selecting a request from the dropdown menu. Then, enter, paste, or upload your code as needed.
- The coding assistant will provide suggestions, completions, and other assistance based on the request prompt you select and the code you provide.
- Feel free to explore the other features of CodeMaxGPT to assist you in your coding tasks.
For a detailed demonstration of using CodeMaxGPT, please visit the introduction page of the web application.
Contributions are welcome! If you would like to contribute to the development of these web applications, please follow these steps:
- Fork the repository.
- Create a new branch for your feature or bug fix.
- Make your changes and commit them.
- Push your changes to your forked repository.
- Submit a pull request detailing your changes.
Please ensure that your contributions align with the project's coding conventions and standards. Your efforts are greatly appreciated!
This project is licensed under the MIT License. Feel free to modify and use the code for your own purposes. However, please note that OpenAI's usage policies and guidelines still apply when using the OpenAI API.
I would like to acknowledge the following organizations and technologies for their contributions to this project:
- OpenAI for developing the powerful language models and APIs, which have enabled me to create these web applications.
- Streamlit for providing a wide range of widgets and Streamlit Cloud, which is a platform that allows me to deploy these web applications easily and efficiently.
- Python for providing a powerful programming language that has been instrumental in the development of these applications.
- GitHub for hosting this repository and providing a collaborative platform for open-source development.
Thank you for choosing to use the OpenAI API Web Apps. I hope that these applications will greatly amplify your programming capabilities and boost your efficiency, both in your work and in your everyday life.