Qwen Code Outlines Privacy Policies Across Four Authentication Methods

The Qwen Code team published its Qwen Code terms of service and privacy notice for the open-source AI coding assistant tool.

The documentation details how privacy rules and terms depend on which authentication method a user selects. Qwen Code supports four authentication methods to access AI models: Qwen OAuth, Alibaba Cloud Coding Plan, bring-your-own API keys, and Google Cloud Vertex AI.

For users logging in with Qwen OAuth, terms and data policies are set by Qwen AI. The documentation notes that the free tier for Qwen OAuth was discontinued on 2026-04-15.

When using the Alibaba Cloud Coding Plan, users follow Alibaba Cloud terms across two regional console endpoints: aliyun.com and alibabacloud.com. Users who authenticate with their own API key or Vertex AI are subject to the policies of their chosen provider or Google Cloud.

Qwen Code itself does not use prompts, code, or responses to train AI models. Any model training depends on the rules of the underlying provider.

The tool collects optional anonymous usage statistics and telemetry data to evaluate product quality and user experience. Collected data includes commands run, performance metrics, crash reports, error reports, and feature usage patterns. Qwen Code explicitly does not collect user code, prompts, model responses, or personal information.

Data collection settings vary by setup. Qwen OAuth and Alibaba Cloud Coding Plan users can opt out of telemetry through Qwen Code settings. For custom API keys and Vertex AI, Qwen Code collects no extra data beyond what those providers collect.

Users can switch between authentication methods during tool startup, inside the CLI using the /auth command, or by setting .env environment variables.