Anyone evaluated the difference between Qwen Code for the local qwen models vs another harness? CC, OC, LC, Aider etc..

A user asks for a comparison between Qwen Code and other harnesses (like opencode) for evaluating local Qwen models. They wonder if Qwen Code offers superior native functionality and what benchmarking methodology was used.