Image Recognition (English)

Author: dugufeng Tags: language:english, vision, multimodal, qwen-vl, siliconflow

This is a straightforward multimodal workflow. It receives an uploaded image and uses the Qwen2.5-VL-32B-Instruct model to return a concise description of the image in English.

---

🛠️ 关键元数据 (Technical Details)

• Dify Version: v1.9.0+ (Please fill in the Dify version you have tested)

---

🚀 关键前置条件 (Pre-conditions)

1. SiliconFlow API Key:

• This workflow relies on the langgenius/siliconflow provider.
• Please ensure you have configured a valid API Key for SiliconFlow in your Dify "Credentials".

🚀 如何使用

1. Set API Keys:
   • Ensure your SiliconFlow API Key is correctly set up in Dify's "Credentials".
2. Configure Tools:
   • No other tools need configuration.
3. Run the Workflow:
   • In the "Start" node's imageUrl variable, upload an image (or provide a URL).
   • Run the workflow and check the "End" node for the text output (the image description).

🛠️ 工作流节点 (Optional)

• Start Node: Receives a file type input named imageUrl.
• LLM (Vision) Node:
  • Uses Qwen/Qwen2.5-VL-32B-Instruct model with vision enabled.
  • Receives the imageUrl as context.
  • Uses an English System Prompt (see above) to generate a concise description.
• End Node: Outputs the final text (string) description.

📸 运行截图 (Highly Recommended)

工作流图

运行示例