China AI Hub
Agents / Qwen Code
Qwen Code
Alibaba Qwen team's open-source AI coding agent (Apache-2.0) for terminal, editor, desktop, browser and chat. Originally based on Google Gemini CLI v0.8.2, independent development since v0.1 as a multi-protocol, multi-platform agent framework. Ships as CLI (npm), Desktop app, VS Code 'Qwen Code Companion' (Beta), Web UI and IM channels (Telegram/DingTalk/WeChat/Feishu). Includes 5 permission modes, Seatbelt/Docker sandboxing, auto-memory, subagents, MCP, computer use and multi-protocol model support.
Qwen Code is a coding agent by alibaba-cloud, deployable self-hosted, and released open source.
| Key fact | Value |
|---|---|
| Company | alibaba-cloud |
| Type | coding |
| Framework | Multi-protocol agent framework (TypeScript); supports OpenAI, Anthropic, Gemini, Qwen APIs plus DeepSeek, MiniMax, Z.AI, Kimi, OpenRouter and local models (Ollama/vLLM) |
| Deployment | self_hosted |
| Open source | Yes (Apache-2.0) |
| GitHub | https://github.com/QwenLM/qwen-code |
| Documentation | https://qwenlm.github.io/qwen-code-docs/en/users/overview/ |
| Pricing | CLI free (Apache-2.0); user pays the model provider. Alibaba Cloud Coding Plan (intl) Pro $50/month; Token Plan (CN, Beijing only) Personal Lite ¥39 / Essential ¥79 / Standard ¥139 / Pro ¥499 per month, team seats ¥150-¥1398; or pay-as-you-go Model Studio API keys; BYO keys to other providers. |
Capabilities
| Capability | Supported |
|---|---|
| Tool calling | Yes |
| Browser use | Yes |
| Computer use | Yes |
| MCP support | Yes |
| Memory | Yes |
| Planning | Yes |
| Multi-agent | Yes |
| API | Yes |
Use cases
- Building features from natural-language descriptions
- Debugging and fixing issues in existing codebases
- CI automation and pipe-friendly Unix workflows (qwen -p)
- Remote agent via chat channels (Telegram, DingTalk, WeChat, Feishu)
- Multi-model head-to-head comparisons via Agent Arena
Limitations
- Web UI and daemon (qwen serve) marked experimental
- Qwen OAuth free tier discontinued 2026-04-15
- Sandboxing reduces but does not eliminate all risks; GUI apps may not work in sandboxes
- Default Docker sandbox image is intentionally minimal (Java not included by default)
- Auto Mode classifier biased toward blocking and fails closed on classifier outage
- Auto-memory is best-effort; QWEN.md is the guaranteed instruction file
Qwen Code is the Qwen team’s open-source AI coding agent (Apache-2.0) for terminal, editor, desktop, browser and chat. Its multi-protocol TypeScript framework supports OpenAI, Anthropic, Gemini and Qwen APIs plus DeepSeek, MiniMax, Z.AI, Kimi, OpenRouter and local models (Ollama/vLLM).
The CLI is free and the user pays the model provider. Alibaba Cloud billing options include the international Coding Plan (Pro $50/month), the China Token Plan (Personal ¥39–¥499/month, team seats ¥150–¥1,398) or pay-as-you-go Model Studio API keys.
See the Alibaba Cloud profile.
Sources
Data interpretation
| Field | Value | Evidence type |
|---|---|---|
| Company | alibaba-cloud | Official |
| Type | coding | Official |
| Open source | Yes (Apache-2.0) | Official |
| Deployment | self_hosted | Official |
| Capabilities | Tool calling, Browser use, Computer use, MCP support, Memory, Planning, Multi-agent, API | Official |
Evidence types: Official = vendor documentation or official pages. See the sourcing policy.
What is Qwen Code?
Alibaba Qwen team's open-source AI coding agent (Apache-2.0) for terminal, editor, desktop, browser and chat. Originally based on Google Gemini CLI v0.8.2, independent development since v0.1 as a multi-protocol, multi-platform agent framework. Ships as CLI (npm), Desktop app, VS Code 'Qwen Code Companion' (Beta), Web UI and IM channels (Telegram/DingTalk/WeChat/Feishu). Includes 5 permission modes, Seatbelt/Docker sandboxing, auto-memory, subagents, MCP, computer use and multi-protocol model support.
Is Qwen Code open source?
Yes — licensed under Apache-2.0.
How much does Qwen Code cost?
CLI free (Apache-2.0); user pays the model provider. Alibaba Cloud Coding Plan (intl) Pro $50/month; Token Plan (CN, Beijing only) Personal Lite ¥39 / Essential ¥79 / Standard ¥139 / Pro ¥499 per month, team seats ¥150-¥1398; or pay-as-you-go Model Studio API keys; BYO keys to other providers.
Where does China AI Hub get its Qwen Code data?
From 4 sources, last verified 2026-09-20.