Leveraging Groq Cloud's inference infrastructure for instant transpilation from pseudo-code to Python.
Translates your .no files into production-ready Python code in milliseconds using Groq's ultra-fast inference.
Write in plain English or simplified pseudo-code. NovaCompile understands intent and generates clean, idiomatic Python.
Only requires the lightweight requests library. No heavy SDKs or complex setup required.
Three simple steps to configure NovaCompile inside your local workspace or web-shell environment.
Clone the repository to get Nova on your device.
cd ~
mkdir .novacompile
cd .novacompile
git clone https://github.com/novacompile/compiler.git
cd compiler
Go to console.groq.com, sign up, and create an API key.
⚠️ Remember to copy the key and put it in a safe and secure place straight away after creating it.
Run the setup script to install dependencies, create a permanent alias, and install your API key.
bash setup.sh
ℹ️
To make this shortcut permanent across terminal restarts, the setup script appends the alias directly into your ~/.bashrc configuration profile.
Once configured, create a custom code file using your preferred pseudo-syntax or simplified structures.
Launch the shell directly to type Nova code interactively:
nova
If you run the transpiler without a file argument, it starts this shell automatically. In shell mode, type a block of Nova code, then press Enter on an empty line to transpile and run it. Use exit or quit to leave.
Create a file with the .no suffix (e.g. script.no) containing your custom logic.
// Define some basic variables
username is "john"
counter is 5
// Run a structural loop
loop 3 times
add 2 to counter
display text "User logged in: " and then show username
print "Current count value is: " + counter
Execute the file instantly by calling nova followed by your custom script target:
nova [filename].no
If you prefer to run the entrypoint directly, this also works (must be inside the root directory of compiler):
python src/transpiler.py [filename].no
⚠️
To be able to run the entrypoint directly, you must be inside the root directory of compiler (by default it is ~/.novacompile/compiler).
Write code the way you think. NovaCompile interprets simplified syntax and translates it into clean, executable Python using Groq's LLM inference.
// Define some basic variables
username is "James"
counter is 5
// Run a structural loop
loop 3 times
add 2 to counter
display text "User logged in: " and then show username
print "Current count value is: " + counterStart writing in Nova syntax today and let AI handle the translation to production Python.