I just published the new Open-source project: ChatGPT Max 3.0.
With ChatGPT Max, you can bypass many traditional ChatGPT limits, and you will be able to do things you never thought about!
Disclaimer: This is not an official edition developed by OpenAI, the company behind ChatGPT. This edition is a customized version I developed after hours of research and testing.
In this post, we will cover the following:
- What is ChatGPT Max?
- ChatGPT Max main features.
- How to set up and run?
- How to customize?
- Limitations & Upcoming Features.
What is ChatGPT Max?
The idea is simple, ChatGPT Max is a customized version of ChatGPT that is built with custom scripting (python) to achieve things you never thought about with traditional ChatGPT, even with GPT-4 or ChatGPT Plus.
The current prototype of ChatGPT Max Allows us to do the following:
- Search Google and Do small research based on live results.
- Rewrite text with one command.
- Summarize blog posts with one command.
- Ability to add more commands (in code)
- Set Custom Models Like GPT-3.5-16K.
- Token and Cost Estimation.
- Set Custom Temperature and Top_p.
I used the latest streamlit chat components to build a simple, clean UI for this ChatBoT.
ChatGPT Max Main Features
As I mentioned, this is the first prototype of version 3.0, and I have a large list of features that will be added soon in the next updates that I will discuss in the last section of this post.
But let me share with you now what are the current features:
1- Builtin Commands
One of the main features that I added is built-in commands, so you can run special tasks like searching Google or summarizing with just one command.
For now, we have 3 commands:
This command will help you search Google for the top 3 results, summarize them, and do a small research about any topic you want. Here is an example of how to use it:
/google email marketing tips
This command will search Google for “email marketing tips” do a small research, and give you the results with the sources.
This command will help you summarize any online blog post within that chat window. Example:
This command will help you rewrite any piece of text in one shot. Try it!
/rewrite your text goes here
2- Model Customization
The second feature is the ability to use any OpenAI Model you want. And not only GPT-3.5 or GPT-4.
You can pick any model like GPT-3.5-16K, which allows working with 16K tokens. This is very important when you work with large inputs or if you want to generate long content.
3- Play with Parameters
With ChatGPT max, you can play with Temperature and Top_p while you are chatting as if you were in the OpenAI playground. This feature is very important, especially if you are into prompt engineering, where we test, evaluate, and craft power prompts.
4- Token Calculation & Estimation
While you are chatting, all tokens and costs will be tracked and updated on the screen. This will help you see how much your chat costs and plan your billing accordingly.
💡Note that this is calculated on a session basis and will reset automatically when you refresh the app.
How to set up ChatGPT Max?
I tried my best to make the project as simple as possible so anyone could download and test it.
Step 1: Download the project
You can download the project directly by clicking on the button below:
or you can clone the repository from GitHub.
Step 2: Get an OpenAI Key
To use any of OpenAI’s models, you must first obtain an API access key.
Go to https://openai.com/api/ and click on the “Signup” button.
You will be prompted to create an account. Verify your email to complete the signup process.
Once logged in, you can find your API keys under the account dashboard.
Be sure to keep your keys private, as anyone with access can use your OpenAI quota.
Now, you can use the key directly in the UI.
Step 3: Install Requirements
I added all the requirements and packages to the requirements .txt file. To install, you need to run this command in the terminal:
pip install -r requirements.txt
And wait a couple of minutes for all the packages to be downloaded and installed.
Step 4: Run the project
now, we are ready to run the project, use this command:
streamlit run chat.py
You can customize the code to help fit more in your scenario.
Note that these customizations are planned to be done later through the UI. But I want to mention it now if you want to try it yourself.
1- Add More Models
In the project, I have added 3 models that you can choose from.
You can find it in the chat.py file here:
You can simply add more models to the list. But make sure then to update the cost calculation method in the tokens_count.py file here:
2- Change Google Search Results Pages
By default, I made the Google Search Function return the top 3 results. You can change this in the google_serp.py file:
Note that increasing this number will make the model take more time to read and summarize pages. You can play with this number and test it.
3- Change Prompts
You will find some prompts I used to summarize and rewrite text in the project. You will find these prompts in the prompts.py file in the project.
Soon, I will be updating these prompts with Power Prompts from my premium prompts library. You can do this yourself in case you already have access to the library.
A Couple of months ago, I started working on a project called “one-shot-gpt”
it is an open-source library that helps facilitate working with language models.
I found that there are many things common between the ChatGPT Max project and the OneShot GPT.
So I decided to merge both in one project and chose the Max to be the name!
This is just the beginning, and I plan over 50 updates and features.
Here is a list of the top planned updates:
- Organizing Chats in Folders and Categories
- Adding Data Analysis Capabilities
- Document Embedding
- Working with Custom and Open Source Language Models.
- Save Costs and track through the whole app.
- Optimize Google Search and Integrate some Serp APIs.
- Add More Built-in commands (Target: 100 Commands)
- Optimize the project code.
And much more! Stay updated by joining my Weekly Newsletter.
🟥Master the Most In-Demand Skill of the Future!
Enroll in the ‘Become a Prompt Engineer‘ program. We’ll take you from novice to expert in scripting AI workflows. Start your journey here!
What Will You Get?
- Access to our premium Prompt Engineering Course
- Access our private support forum to get help along your journey.
- Access to our Premium Tested Prompts Library.