{
  "id": 12689714,
  "title": "Touch Grass, Pack Your Docs: An Offline Coding Agent on a 4 GB Laptop GPU",
  "url": "https://urgent.news/2026/10/07/touch-grass-pack-your-docs-an-offline-coding-agent-on-a-4-gb-laptop",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-07T18:55:15.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/gde/touch-grass-pack-your-docs-an-offline-coding-agent-on-a-4-gb-laptop-gpu-4lf1"
  },
  "original_language": "en",
  "account": "This story details the development of an offline coding agent that can write working code on a laptop with a 4 GB GPU, even without internet access. The Gemma 4 E4B model is used, running locally on a GTX 1650 Ti GPU, with the llama.cpp inference engine and opencode agent. Six benchmark runs using the agent achieved full pass rates, while other combinations of components failed 16 out of 16 tests. The agent is designed to work in low-connectivity environments, like parks, campsites, or trains, by packing necessary documents, model, server, and agent on the laptop itself.",
  "summary": "This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass This article provides a step by step build of an offline coding agent on a laptop with a 4 GB GPU. Gemma 4 runs locally under llama.cpp, opencode drives it, and every run is scored by an independent grader, from the first context sweep to the setup that passes. https://github.com/xbill9/dev-hacktoberfest Gemma…",
  "key_points": [
    "Offline coding agent developed for 4 GB laptop GPU",
    "Gemma 4 E4B model runs locally on GTX 1650 Ti GPU",
    "Agent achieves full pass rates in six benchmark runs"
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}