🔥 This Tool Picks the PERFECT Local LLM for Your GPU

Stop wasting time downloading LLMs that don't fit your GPU or run too slow. whichllm auto-detects your hardware (NVIDIA, AMD, Apple Silicon, CPU) and ranks models by REAL benchmarks — not just parameter count.

One command: uvx whichllm@latest

It pulls live data from HuggingFace, merges 6 benchmark sources (LiveBench, Chatbot Arena, Aider coding tests, vision evals), and shows you the top models that actually run well on YOUR system. You can even simulate GPUs before buying: whichllm --gpu "RTX 5090"

Over 2,700 stars because it solves the "which model should I download?" problem developers face every single day. Built by @Andyyyy64, MIT licensed, Python — installs in seconds with uv or pip.

Repo: https://github.com/Andyyyy64/whichllm

Видео 🔥 This Tool Picks the PERFECT Local LLM for Your GPU канала Noborta