Worth Doing AI

We provide high-quality GGUF quantizations of the best open-source language models, optimized for local inference on Apple Silicon Macs.

What We Do

We select the best small general-purpose models and quantize them using llama.cpp with carefully chosen quantization levels. Every model is tested on Apple Silicon hardware before release.

Our focus:

• Best-in-class small models (1.7B to 7B parameters)
• GGUF format compatible with Ollama, LM Studio, and llama.cpp
• Optimized for Apple Silicon (Metal GPU acceleration)
• Multiple quantization levels to fit any hardware

Available Models

Quantization Levels

Quick Start

Ollama

# Download a GGUF file, then:
cat > Modelfile <<'EOF'
FROM ./qwen2.5-7b-instruct-Q4_K_M-worthdoing.gguf
EOF
ollama create qwen2.5-7b -f Modelfile
ollama run qwen2.5-7b

llama.cpp

llama-cli -m qwen2.5-7b-instruct-Q4_K_M-worthdoing.gguf -p "Your prompt" -ngl 99

LM Studio

Download any GGUF file and import it directly into LM Studio.

Hardware Recommendations

About

Worth Doing AI is focused on making high-quality AI accessible for local, private use. All quantizations are performed with llama.cpp and verified on Apple Silicon hardware.

Contact: admin@worthdoing.ai