LLMFit
ActiveDescription
Hundreds of models and providers. One command to find what runs on your hardware. Provides local LLM runtime for AI agents.
Key Features
- Auto hardware detection & model matching — Detects CPU, RAM, GPU/VRAM, scores hundreds of models across quality, speed, fit, and context dimensions
- Interactive TUI interface — Default terminal UI with search, sort, multi-select comparison, hardware simulation, community leaderboard
- Multi-GPU & MoE support — Supports multi-GPU setups and Mixture-of-Experts architecture model evaluation
- Dynamic quantization selection — Auto-recommends optimal quantization per hardware, shows memory usage and run mode
- Runtime provider integration — Supports Ollama, llama.cpp, MLX, Docker Model Runner, LM Studio and other local runtimes
- Download manager — Built-in model download, history tracking, deletion, and configurable download directory
Use Cases
💡 Local LLM selection — Evaluate which models run smoothly on your hardware before purchasing
💡 AI Agent local inference deployment — Select appropriate local inference models and assess performance for AI agents
💡 Hardware configuration planning — Simulate model performance across different hardware configurations
💡 Model comparison testing — Multi-select models to compare tok/s, TTFT, VRAM usage side by side
💡 Team selection collaboration — View real-world performance data from other users with the same hardware via community leaderboard
Categories
Quick Start
brew install AlexsJones/llmfit/llmfit
# Launch interactive TUI
llmfit
# Or use CLI mode for recommendations
llmfit recommend --use-case coding
# Run via Docker
podman run ghcr.io/alexsjones/llmfit recommend --use-case coding | jq '.models[].name'