macOS Guide

python3 -m venv .venv
source .venv/bin/activate
pip install 'nexaai[mlx]'

from nexaai import LLM, ModelConfig
llm = LLM.from_(model="NexaAI/Qwen3-1.7B-4bit-MLX", config=ModelConfig())
for t in llm.generate_stream("Hello", None):
    print(t, end="")