You dropped 2k on a MacBook just to browse Reddit? Check out Apfel, an open-source tool that exposes the native, free AI baked right into macOS.
You dropped two grand on a shiny M-series MacBook, and I bet 90% of you are just using it to write console.log and doomscroll Reddit. Truth is, Apple secretly baked a ton of AI capabilities into macOS ages ago, but most of us didn't know how to unleash it.
Enter Apfel, an open-source project by some absolute madlad named Arthur-Ficial. It literally exposes the hidden AI inside your Mac, and it’s completely free. The post just hit Hacker News and casually bagged over 544 upvotes.
Apfel (German for Apple, clever guy) isn’t some bloated 40GB LLM you have to torrent and run until your laptop melts. TL;DR for the lazy: it’s a super lightweight wrapper that hooks directly into the native Machine Learning frameworks Apple already sneaked into macOS.
Instead of installing RAM-hungry AI tools wrapped in dirty Electron, or donating your monthly salary to Sam Altman for OpenAI API keys, Apfel uses what you already paid for. It runs 100% locally. No internet required. No leaking your spaghetti code to third-party servers.
With 500+ upvotes, the pragmatic dev community on HN is eating this up. Based on the massive upvote ratio, you can spot the usual suspects cheering in the background:
We devs often get blinded by massive, hype-train tech trends and forget that the machines under our wrists are stupidly powerful. Apfel isn't going to replace Claude 3.5 Sonnet or ChatGPT for complex architectural design, but it’s a brilliant utility for quick, secure, local tasks.
Give it a spin. It’s open-source on GitHub. Worst case, you waste a few megabytes of storage.
Source: Hacker News | Github: Arthur-Ficial/apfel
