How to integrate 通义千问 (Qwen) in Higress AI Gateway

通义千问是阿里云自主研发的超大规模语言模型，具备强大的语义理解、内容创作、多轮对话和逻辑推理能力。

Core Advantages

原生中文支持优化

多模态理解能力

支持长文本输入

内置联网搜索增强

provider:
  type: qwen
  apiTokens:
    - "YOUR_QWEN_API_TOKEN"
  modelMapping:
    "gpt-3": "qwen-turbo"
    "gpt-4": "qwen-max"

Note: Please replace YOUR_API_TOKEN with the actual token you applied for on the 通义千问 (Qwen) platform.

By accessing 通义千问 (Qwen) through Higress AI Gateway, you not only get simple protocol forwarding, but also enterprise-level governance capabilities:

Unified Protocol Adaptation: Use the standard OpenAI protocol to call 通义千问 (Qwen) without worrying about underlying differences.
Intelligent Retry and Circuit Breaking: When the model service is unstable, Higress automatically handles retries to ensure business continuity.
Security Audit and Rate Limiting: Protect your API Key from abuse and perform precise rate limiting based on Token consumption.
Semantic Caching: Significantly reduce response time for repeated questions and save inference costs.