← All models

Qwen-Doc-Turbo

Published Publisher Alibabaqwen-doc-turbo

Model Positioning

Qwen-Doc-Turbo is a text generation model for document data mining, designed specifically for information extraction, content moderation, classification and tagging, and summary generation. The date listed in the official release record is 2025-07-23. (docs.modelstudio.console.alibabacloud.com)

Capabilities and Uses

  • Extracts fields from documents and can generate structured text such as JSON based on prompts, making it suitable for document archiving, information organization, content moderation, and other scenarios.
  • Supports document processing through publicly accessible file URLs, uploaded file IDs, or directly supplied plain text. The domestic Alibaba Cloud Model Studio documentation also describes generating PPT presentations from documents. (help.aliyun.com)

Limitations

  • Official documentation states that the model is available in the North China 2 (Beijing) region and requires an API Key for that region. The file URL method supports only the DashScope protocol and allows up to 10 publicly accessible file URLs per request; the file ID method allows at most one file per request; directly supplied plain text is limited to 9,000 tokens. (help.aliyun.com)
  • The official model information page lists a context length of 262,144 tokens, a maximum input of 253,952 tokens, and a maximum output of 8,192 tokens; however, the pricing table in the data mining guide lists the maximum output as 32,768 tokens. The official pages differ, so the limits of the platform in use and the API response should be taken as authoritative before making an actual call. (help.aliyun.cn)
  • The model information page states that Function Calling, web search, and dedicated structured output functionality are not supported. Generating JSON text based on prompts is not equivalent to supporting a structured output API that guarantees the format. (help.aliyun.cn)

Pricing

All prices below are officially published token-based prices. Currency quotations from different platforms are not treated as the same package that can be directly converted; all were checked on 2026-11-18.

  • Domestic site prices in CNY: For Alibaba Cloud Model Studio in the North China 2 (Beijing) region, standard input is CNY 0.6 per million tokens and output is CNY 1 per million tokens; an input cache hit is CNY 0.12 per million tokens, explicit cache creation is CNY 0.75 per million tokens, and an explicit cache hit is CNY 0.06 per million tokens. The model information page states that the prices shown are the original API usage prices and exclude limited-time discounts. Source: Alibaba Cloud Model Studio model information and pricing. (help.aliyun.cn)
  • International site prices in USD: The Alibaba Cloud Model Studio documentation lists input at USD 0.087 per million tokens and output at USD 0.144 per million tokens, with no free quota; the documentation applies to the North China 2 (Beijing) region. Source: Alibaba Cloud Model Studio data mining documentation and pricing. (alibabacloud.com)

Official Sources