Qwen: comments stay in the language the file already uses.

Alibaba Qwen is seated for bilingual coding. Spoken Mandarin stays with DeepSeek. Dossiers stay with Kimi. This rail writes the repo — 中英 mixed, without flipping the file.

Qwen
Qwen-Max · Qwen-Plus · Qwen-Turbo
Tongyi Qianwen, bilingual code, 中英 repos
Qwen Code

Tongyi Qianwen, then a coder’s house

Qwen is Alibaba’s model family, born as Tongyi Qianwen and trained to be natively bilingual. Qwen2 and Qwen2.5 made the line a default choice for 中英 work. Qwen-Max is the flagship. Qwen-Plus is the everyday seat. Qwen-Turbo is the fast one. Qwen-Coder taught the models to live in a repository instead of a chat box. On Wrooms that last skill is the whole rail.

Max, Plus, and Turbo

Qwen-Max is the deep seat for a hard multi-file change. Qwen-Plus handles everyday edits. Qwen-Turbo is the short flash. You do not type those names in the room. You @qwencode, paste the file, and say what should move. Comments stay in the language the file already uses.

What Qwen Code will not become

A Mandarin chat partner — that is DeepSeek. A sourced 中文 brief — that is Kimi. A form extract — that is GLM. A JSON tool trace — that is Hermes. An outage — that is house Fixes.

How to work a mixed repo

Paste the file and name the function. @qwencode. Pin when the room is a coding thread and should not become a research crawl.