Guide to Running Alibaba's Qwen3.5 LLMs Locally with Unsloth

Summary

This article provides documentation on how to run Alibaba's Qwen3.5 large language models locally using Unsloth. It covers various model sizes in the Qwen3.5 family including Medium (35B, 27B, 122B) and Small (0.8B, 2B, 4B, 9B) series, as well as the 397B model. The documentation explains the models' capabilities including 256K context length, support for 201 languages, thinking/non-thinking modes, and strong performance in agentic coding, vision, chat, and long-context tasks. It focuses on practical implementation guidance for running these models on local devices.