Qwen Code

Spec sheet

Purpose

Qwen Code is an open-source AI coding agent available through terminal, editor, desktop, browser, and chat interfaces.Read 7 Oct 2026 · github.com

Platforms

The official repository offers desktop downloads for macOS, Windows, and Linux, a browser Web UI, and a VS Code extension.Read 7 Oct 2026 · github.com

Integrations

The overview describes integrations with Google Drive, Figma, Slack, and Jira through MCP, and the documentation lists messaging channels including Telegram, WeChat, QQ, DingTalk, WeCom, and Feishu.Read 7 Oct 2026 · qwenlm.github.io

Privacy

Optional usage statistics may include command, performance, error, and feature data; Qwen Code says it does not collect code, prompts, or model responses as usage statistics.Read 7 Oct 2026 · qwenlm.github.io

Support

The documentation directs users to the Qwen Code issue tracker for questions and feedback.Read 7 Oct 2026 · qwenlm.github.io

Agent features

Its listed capabilities include subagents, agent teams, dynamic workflows, automatic memory and skills, hooks, code review, plan mode, and sandboxing.Read 7 Oct 2026 · github.com

Authentication availability

The documentation says the Qwen OAuth free tier was discontinued on 2026-04-15 and recommends using a currently supported authentication provider for new setups.Read 7 Oct 2026 · qwenlm.github.io

Automation

Qwen Code can run in headless mode for CI/CD workflows, including GitHub Actions tasks such as replying to issues, reviewing code, and submitting pull requests.Read 7 Oct 2026 · qwenlm.github.io

Browser support

Browser Use requires macOS or Linux, Chrome 125 or later, Qwen Code 0.24.2 or later, and a separately installed Chrome extension.Read 7 Oct 2026 · qwenlm.github.io

Coding tasks

It can read and modify local files, run shell commands, fetch web content, and use those results to answer coding requests.Read 7 Oct 2026 · qwenlm.github.io

Editor and chat integrations

The project lists integrations for VS Code, JetBrains, Zed, Telegram, DingTalk, WeChat, and Feishu.Read 7 Oct 2026 · github.com

Extensions

Extensions can package prompts, MCP servers, subagents, skills, and custom commands, and the docs say compatible extensions can be installed from several other ecosystems.Read 7 Oct 2026 · qwenlm.github.io

IDE support

The IDE integration documentation supports Visual Studio Code and editors that support VS Code extensions, with workspace context and native diff review.Read 7 Oct 2026 · qwenlm.github.io

Installation requirement

The npm installation requires Node.js 22 or later.Read 7 Oct 2026 · github.com

Installation requirements

The overview provides installers for Windows, macOS, and Linux and says the npm fallback requires Node.js 22 or later.Read 7 Oct 2026 · qwenlm.github.io

License and pricing

The repository identifies the project as open source and lists an Apache-2.0 license; the opened pages did not state a Qwen Code subscription price.Read 7 Oct 2026 · github.com

MCP integrations

It connects to external tools, databases, APIs, and workflows through MCP using local stdio, SSE, or streamable HTTP transports.Read 7 Oct 2026 · qwenlm.github.io

Model providers

It supports OpenAI-compatible providers, Anthropic, Google Gemini, Vertex AI, Alibaba Cloud, and custom or self-hosted endpoints, with model switching at runtime.Read 7 Oct 2026 · qwenlm.github.io

Pricing

The product overview describes Qwen Code as open-source and free, but the documentation does not list a current product plan or price.Read 7 Oct 2026 · qwenlm.github.io

Privacy controls

Usage statistics are enabled by default in the documented settings and can be disabled through settings or an environment variable.Read 7 Oct 2026 · qwenlm.github.io

Provider data handling

Qwen Code says it does not use prompts, code, or responses for model training; data handling depends on the selected AI provider’s policies.Read 7 Oct 2026 · qwenlm.github.io

Safety controls

Qwen Code typically requests confirmation before sensitive file or shell actions and supports sandboxing tool execution.Read 7 Oct 2026 · qwenlm.github.io

Sandbox limits

The Linux sandbox documentation warns that Landlock provides partial enforcement and cannot enforce closed networking or hide explicit masked paths.Read 7 Oct 2026 · qwenlm.github.io

Security controls

The tool documentation says potentially sensitive actions typically require confirmation and describes sandboxing for tool execution.Read 7 Oct 2026 · qwenlm.github.io

Where Specifiction places it

• AI Code Generators#48 of 547.5
• AI Code Generation Software#24 of 277.5