<oembed><type>rich</type><version>1.0</version><title>Edgaras wrote</title><author_name>Edgaras (npub179…22v2r)</author_name><author_url>https://yabu.me/npub179nsvm5fvtrjsgfk0m9hlz6arqk2eyyw7xk58zm46fapgx327nas722v2r</author_url><provider_name>njump</provider_name><provider_url>https://yabu.me</provider_url><html>You don’t need to drop big money on hardware to run a solid local agent.&#xA;Just picked up a used Lenovo ThinkCentre M715q (Ryzen 5 PRO) on eBay for ~$75, threw in extra 8GB RAM stick for ~$25, installed latest headless Fedora Server, and now running hermes-agent smoothly connected to my local inference server (Halo Strix).&#xA;Low power, quiet, and more than enough.&#xA;No fancy GPUs required.&#xA;Local AI is more accessible than you think!</html></oembed>