Spec2LLM

Find the best LLMs for your hardware specs. CLI tool that detects your system (CPU, GPU, RAM, storage) and recommends compatible models ranked by performance fit.

pip install spec2llm
spec2llm recommend

Features

• Hardware Detection — CPU, GPU (NVIDIA/AMD/Apple/Intel), RAM, storage, OS — cross-platform
• Smart Recommendations — Models ranked by VRAM headroom, RAM availability, GPU compute tier, and CPU cores
• Curated Catalog — 40+ popular models (Llama 3, Mistral, Gemma, Qwen, DeepSeek, Phi, etc.) with tested requirements
• Auto-Discovery — catalog update fetches new models from Ollama registry with estimated requirements
• Apple Silicon Support — Detects unified memory and adjusts scoring accordingly
• JSON Output — --json flag on all commands for scripting/automation

Commands

spec2llm scan              Detect system hardware specs
spec2llm recommend         Find best LLMs for your system
spec2llm search <query>    Search model catalog
spec2llm list              Browse all models
spec2llm install <model>   Show install command (Ollama/HF)
spec2llm compare <a> <b>   Side-by-side comparison
spec2llm catalog update    Discover new models from Ollama

Options

Most commands support:

• --json — machine-readable JSON output
• --top N — limit recommendations (recommend)
• --tag — filter by tag like code, vision, chat
• --run — execute install command directly (install)

Examples

# Quick recommendation
spec2llm recommend

# Filter by use case
spec2llm recommend --tag code --top 5

# Search for models
spec2llm search deepseek

# Compare two models
spec2llm compare llama-3.2-1b-q4 mistral-7b-q4

# See install command
spec2llm install llama-3.1-8b-q4

# Machine-readable output
spec2llm scan --json

How It Works

1. Scan detects your CPU cores/freq, RAM total/available, GPU model/VRAM, free storage, and OS
2. Match filters models that fit your VRAM, RAM, and storage
3. Score (0-100): 40% VRAM headroom, 20% RAM headroom, 20% GPU tier match, 10% CPU cores, 10% Apple Silicon bonus
4. Recommend returns sorted list with scores and details

Installation

pip install spec2llm

Or from source:

git clone https://github.com/Techikrish/Spec2LLM.git
cd Spec2LLM
pip install -e .

Requires Python 3.9+.

New Models

When a new model is released:

spec2llm catalog update

This fetches available models from Ollama's registry and estimates their hardware requirements. Run spec2llm recommend to see if they fit your system.

License

MIT