Hey everyone,
A while ago, we realized a major annoyance: whenever you actually need an AI to summarize a document, write some quick code, or just brainstorm, you're usually on a flight, on the subway, or dealing with terrible cell reception. And bam, ChatGPT won't connect. Plus, there's the growing privacy concern of feeding all your personal data to cloud servers.
So, my team and I started tinkering with a question: "What if we just run the AI directly on the phone's hardware?"
We've been spending our evenings and weekends for months trying to make this work smoothly, and the result is Cortex AI.
The logic is super simple:
You download a highly optimized, small-scale local model (from our library) straight to your device.
Put your phone in airplane mode, go off the grid—the AI replies entirely locally.
Zero data leaves your phone. 100% private.
Some real-world use cases we built this for:
Coding help or summarizing offline docs while on a long flight.
Getting quick answers while traveling abroad without an expensive data roaming plan.
Brainstorming private ideas you just don't want OpenAI or Google to scrape.
Note: We do have an optional "Online Mode" if you want to connect to massive models like GPT-4 or Claude, but the local offline models are completely free, and that's what we really want to test right now.
We're currently trying to gather real user experiences on the local execution side. I'm not here to just spam a link and grab cash; we genuinely want to improve the offline mobile AI space.
If anyone frequently travels, camps, or just loves local LLMs, we'd be super grateful if you could test it out. Brutally honest feedback like "runs too slow on my device," "needs X feature," or "this part of the UI makes no sense" is exactly what we need right now :)
[link] [comments]