LLM Model Selector Pro

A high-performance Rust utility designed to optimize Large Language Model (LLM) selection based on local hardware capabilities. It analyzes your system configuration (RAM, VRAM, CPU) and identifies the optimal parameter sizes for models from both the Ollama Library and the Hugging Face Hub.

Core Capabilities

• Dual-Source Analysis: Simultaneously evaluates registries from Ollama and Hugging Face.
• Hardware-Aware Selection: Detects OS, CPU topology, available RAM, and NVIDIA GPU VRAM.
• Apple Silicon Optimization: Specifically tailored for Mac unified memory architectures.
• Parameter Optimization:
  • Local (Ollama): Identifies precise parameter counts (e.g., 8b, 7b, 30b) that will run natively on your current VRAM/RAM profile.
  • Cloud (Hugging Face): Extracts architectural metadata (Billion parameters, Million parameters) from the community hub to guide selection.
• Categorized Intelligence: Organized discovery across Embedding, Vision, Coding, and Reasoning categories.
• Premium Interface: Uses dynamic progress indicators and formatted tables for a professional terminal experience.

🚀 Simple Start Guide (For Everyone)

If you aren't a developer, don't worry! Follow these 3 simple steps to get your recommendations:

1. Install the "Engine" (Rust):
   • Go to rustup.rs and follow the simple instruction for your computer. This allows your computer to run this tool.
2. Download & Setup:
   • Open your Terminal (on Mac: Press Cmd + Space and type "Terminal").
   • Type this (one line at a time):

     cd Desktop
     git clone https://github.com/GoWtEm/llm-model-selector
     cd llm-model-selector

3. Run & Get Results:
   • Copy and paste this final command:

     cargo run --release

   • Wait a few seconds, and it will show you exactly which models (like "8b" or "70b") are perfect for your specific computer!

---

Technical Architecture

The project is structured for modularity and performance:

• src/hardware.rs: Low-level system resource detection and telemetry.
• src/registry.rs: Asynchronous service for querying model metadata from global registries.
• src/analyzer.rs: The suitability engine that matches model requirements against detected silicon.
• src/main.rs: Application orchestration and user interface.

Prerequisites

• Rust Environment: Install Rust (edition 2024).
• Network Access: For real-time registry synchronization.
• Ollama (Optional): Recommended for local execution of suggested models.

Deployment

1. Build the binary:

   cargo build --release

2. Run Execution:

   ./target/release/llm-model-selector
   # Or via cargo
   cargo run --release

Registry Details

The utility tracks:

1. Ollama Library: Popular and newest models optimized for local quantization.
2. Hugging Face Hub: Trending architectural breakthroughs across the open-source community.

Support

If you find this utility useful, consider supporting its development:

License

This project is licensed under the MIT License - see the LICENSE file for details.