llmfit Hardware Model Matcher Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. llmfit detects your system's RAM, CPU, and GPU then scores hundreds of LLM models across quality, speed, fit, and context dimensions — telling you exactly which models will run well on your hardware. It ships with an interactive TUI and a CLI, supports multi GPU, MoE architectures, dynamic quantization, and local runtime providers (Ollama, llama.cpp, MLX, Docker Model Runner). Installation macOS / Linux (Homebrew) Quick install script Windows (Scoop) Docker / Podman From source (Rust) Core Concepts Fit tiers : perfect (runs great), good (runs well), marginal (runs but tight), too tight (won't run) Scoring dimensions : quality, speed (tok/s estimate), fit (memory headroom), context capacity Run modes : GPU, CPU+GPU offload, CPU only, MoE Quantization : automatically selects best quant (e.g. Q4 K M, Q5 K S, mlx 4bit) for your hardware Providers : Ollama, llama.cpp, MLX, Docker Model Runner Key Commands Launch Interactive TUI CLI Table Output Show System Hardware Detection List All Models Search Models Fit Analysis Model Detail Recommendations Hardware Planning (invert: what hardware do I need?) REST API Server (for cluster scheduling) Hardware Overrides When autodetection fails (VMs, broken nvidia smi, passthrough setups): Accepted suffixes: G / GB / GiB , M / MB / MiB , T / TB / TiB (case insensitive). Context Length Cap REST API Reference Start the server: Endpoints Query Parameters for /models and /models/top Param Values Description limit / n integer Max rows returned min fit perfect\ good\ marginal\ too tight Minimum fit tier perfect true\ false Force perfect only runtime any\ mlx\ llamacpp Filter by runtime use case general\ coding\ reasoning\ chat\ multimodal\ embedding Use case filter provider string Substring match on provider search string Free text across name/provider/size/use case sort score\ tps\ params\ mem\ ctx\ date\ use case Sort column include too tight true\ false Include non runnable models max context integer Per request context cap Scripting & Automation Examples Bash: Get top coding models as JSON Bash: Check if a specific model fits Bash: Auto pull top Ollama model Python: Query the REST API Python: Hardware aware model selector for agents Docker Compose: Node scheduler pattern TUI Key Reference Key Action ↑ / ↓ or j / k Navigate models / Search (name, provider, params, use case) Esc / Enter Exit search Ctrl U Clear search f Cycle fit filter: All → Runnable → Perfect → Good → Marginal a Cycle availability: All → GGUF Avail → Installed s Cycle sort: Score → Params → Mem% → Ctx → Date → Use Case t Cycle color theme (auto saved) v Visual mode (multi select for comparison) V Select mode (column based filtering) p Plan mode (what hardware needed for this model?) P Provider filter popup U Use case filter popup C Capability filter popup m Mark model for comparison c Compare view (marked vs selected) d Download model (via detected runtime) r Refresh installed models from runtimes Enter Toggle detail view g / G Jump to top/bottom q Quit Themes t cycles: Default → Dracula → Solarized → Nord → Monokai → Gruvbox Theme saved to ~/.config/llmfit/theme GPU Detection Details GPU Vendor Detection Method NVIDIA nvidia smi (multi GPU, aggregates VRAM) AMD rocm smi Intel Arc sysfs (discrete) / lspci (integrated) Apple Silicon system profiler (unified memory = VRAM) Ascend npu smi Common Patterns "What can I run on my 16GB M2 Mac?" "I have a 3090 (24GB VRAM), what coding models fit?" "Can Llama 70B run on my machine?" "Show me only models already installed in Ollama" "Script: find best model and start Ollama" "API: poll node capabilities for cluster scheduler" Troubleshooting GPU not detected / wrong VRAM reported nvidia smi not found but you have an NVIDIA GPU Models show as too tight but you have enough RAM REST API: test endpoints Apple Silicon: VRAM shows as system RAM (expected) Context length environment variable

llmfit-hardware-model-matcher