Nazar Kvartalnyi

WhiteLightning - LLM Distillation Tool

by
Build fast, compact AI classifiers with LLM prompts. WhiteLightning automatically generates data, trains a small model (<1MB), and deploys anywhere, including mobile, web, or embedded directly from your terminal.

Add a comment

Replies

Best
Nazar Kvartalnyi
Hey Product Hunt 👋 As a developer, I often found AI tools to be either too bulky or reliant on expensive cloud APIs. That's why we built WhiteLightning, an open-source CLI tool that lets you create ultra-compact, offline AI text classifiers with just one command. Whether you're adding sentiment analysis to a desktop app, deploying on a Raspberry Pi, or classifying support tickets locally, WhiteLightning makes it easy to: • Generate synthetic training data without real datasets • Train models under 1MB in size • Deploy anywhere—from mobile apps to microcontrollers • Maintain full control over your data Built for developers who prefer to own their models and run them locally, WhiteLightning is 100% open-source under GPL-3.0, with classifiers you can use commercially under MIT. I'm excited to see how you might use it. Feel free to ask questions or share your thoughts. I’m here to chat!
Volodymyr Melnyk

What a fantastic tool! Thank you for making it!

Viacheslav Kolos

This is a real game-changer for edge AI! WhiteLightning lets you build ultra-lightweight classifiers straight from LLM prompts — no manual data collection needed. Perfect for mobile and embedded use cases. Amazing to see how far low-code AI has come!

Ruslan P

Thanks god I’m not supposed to make up and code it each time by myself anymore. Brilliant tool!

Dmytro Barchuk

Def worth to try

Oleksandr Voitko

That’s an awesome job, guys!

Vladyslav Vasyliev

Super easy way to build edge AI models from prompts. Works great even on mobile. Really useful tool!

Olha Siniatnikova

Tried WhiteLightning and wow — super smooth experience and feels really intuitive.
Definitely a tool I’ll be keeping an eye on!

Natalia Zamai

This looks super cool! Excited to see what folks build with it

svitlanamartyniuk

Definitely curious to see how it evolves 👀