Every project runs on a normal laptop with no GPU, calling models through a hosted inference provider (Nebius or OpenRouter), configured once at the repo root. Some have other local requirements, see ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results