图像理解

ChatGLM / GLM-V 视觉模型适合图片问答、截图分析、票据识别、表格和图表理解。

图片输入

{
  "model": "glm-5v",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "Analyze the main content of this image"
        },
        {
          "type": "image_url",
          "image_url": {
            "url": "https://example.com/image.png"
          }
        }
      ]
    }
  ]
}

截图分析

Analyze this admin screenshot. Identify the current page, key data, abnormal states, and next steps.

票据和图表

Extract merchant name, date, amount, and line items from the image. Use null for unreadable fields and do not guess.

最佳实践

• 图片上传前裁剪无关区域
• 票据和表格不要过度压缩
• 关键金额、日期、证件号要二次校验
• 不要让视觉模型单独承担高风险审核决策