Qwen has released Qwen3-4B-Instruct-2507 & Qwen3-4B-Thinking-2507 – smarter, sharper, and ready for 256K context!

These tiny 4B models punch way above their weight in reasoning, multilingual tasks, math, code, and more. And we can run them locally on Mac in minutes!

Just follow these quick steps:
1 – Download & Install LM Studio
2 – In the search tab type “Qwen3-4B-Instruct-2507” or, if you prefer the Thinking model, “Qwen3-4B-Thinking-2507“
3 – Hit download and start chatting