Project Files
docs / qwen_code_v4.md
qwen code at home.This guide explains how to set up the Qwen Code (v0.14.x+) with local models running via LM Studio. (the Least v0.15.9 -- may 2026)
npm install -g @qwen-code/qwen-code@latest
http://localhost:1234).16k++. The agent's system prompt is very large; smaller contexts will cause the model to "forget" its tools.The qwen code is extremely strict about the JSON schema.
If it's incorrect, the app will rename the file to settings.json.corrupted.
Create/edit the file at:
%USERPROFILE%\.qwen\settings.json~/.qwen/settings.jsonUse this exact structure:
The file settings.json in utf-8.
Qwen Code is the CLI agent.Qwen Code Companion is the pugin for VS Code.This is a VS Code plugin developed by Alibaba Group. It's an intelligent assistant based on the Large Language Model (LLM) that helps you write and debug code right in the editor. Even more settings https://github.com/QwenLM/qwen-code also see authentication guide https://qwenlm.github.io/qwen-code-docs/en/users/configuration/model-providers/ This link is provided by qwen code itself. There is outdated information there.
At 15.04.2026 qwen code requires API key from Alibaba. Here is the instruction I tested.
Qwen Code via LM Studio old version of instruct.
qwen3.5-9b-sushi-coder-rl is preferred LLM model for plugins TS -- IMHO, from my list.txt of LLM.0.0 or 0.1 in LM Studio for more predictable, structured code output.
{
"provider": {
"openai": {
"apiKey": "sk-local",
"baseURL": "http://localhost:1234/v1",
"defaultModel": "qwen3.5-9b-sushi-coder-rl"
}
},
"security": {
"auth": {
"selectedType": "openai",
"apiKey": "sk-local",
"baseUrl": "http://localhost:1234/v1"
}
},
"model": {
"providerId": "openai",
"name": "qwen3.5-9b-sushi-coder-rl"
},
"$version": 4
}