基础地址:https://www.moyu.info

接口:POST /v1/messages

认证:x-api-key: sk-xxx

必须请求头:anthropic-version: 2023-06-01

---

1. 支持的模型

以下模型均已通过实际测试,全部支持图片/视觉输入

模型 ID

系列

测试结果

claude-opus-4-6

Claude 4.6 Opus

通过

claude-sonnet-4-6

Claude 4.6 Sonnet

通过

claude-opus-4-5-20251101

Claude 4.5 Opus

通过

claude-sonnet-4-5-20250929

Claude 4.5 Sonnet

通过

---

2. 图片输入格式

messages[].content 中,图片作为 type: "image" 的内容块传入,支持两种 source 类型:

2.1 Base64 编码(推荐)
{
  "type": "image",
  "source": {
    "type": "base64",
    "media_type": "image/png",
    "data": "<base64编码的图片数据>"
  }
}
2.2 URL 引用
{
  "type": "image",
  "source": {
    "type": "url",
    "url": "https://example.com/image.png"
  }
}

注意:URL 方式要求图片地址可被服务器直接访问(无需登录、无防盗链),否则会报 403 错误。推荐使用 base64 方式确保稳定性。

2.3 支持的图片格式

media_type

格式

image/png

PNG

image/jpeg

JPEG

image/gif

GIF

image/webp

WebP

---

3. 完整请求示例

3.1 Base64 图片输入(单张图片)
curl -X POST "https://www.moyu.info/v1/messages" \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-你的密钥" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "image",
            "source": {
              "type": "base64",
              "media_type": "image/png",
              "data": "iVBORw0KGgoAAAANSUhEUgAAAAo...(base64数据)"
            }
          },
          {
            "type": "text",
            "text": "请描述这张图片的内容。"
          }
        ]
      }
    ]
  }'
3.2 URL 图片输入
curl -X POST "https://www.moyu.info/v1/messages" \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-你的密钥" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "image",
            "source": {
              "type": "url",
              "url": "https://httpbin.org/image/png"
            }
          },
          {
            "type": "text",
            "text": "请描述这张图片的内容。"
          }
        ]
      }
    ]
  }'
3.3 多张图片输入

一次请求中可以传入多张图片,支持 base64 和 URL 混合使用:

curl -X POST "https://www.moyu.info/v1/messages" \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-你的密钥" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "image",
            "source": {
              "type": "base64",
              "media_type": "image/png",
              "data": "iVBORw0KGgoAAAANSUhEUg...(第一张图base64)"
            }
          },
          {
            "type": "image",
            "source": {
              "type": "url",
              "url": "https://httpbin.org/image/png"
            }
          },
          {
            "type": "text",
            "text": "请分别描述这两张图片的内容。"
          }
        ]
      }
    ]
  }'
3.4 Python 示例(本地图片文件)
import base64
import requests

# 读取本地图片并转为 base64
with open("my_image.png", "rb") as f:
    img_base64 = base64.b64encode(f.read()).decode()

response = requests.post(
    "https://www.moyu.info/v1/messages",
    headers={
        "Content-Type": "application/json",
        "x-api-key": "sk-你的密钥",
        "anthropic-version": "2023-06-01",
    },
    json={
        "model": "claude-sonnet-4-6",
        "max_tokens": 1024,
        "messages": [
            {
                "role": "user",
                "content": [
                    {
                        "type": "image",
                        "source": {
                            "type": "base64",
                            "media_type": "image/png",
                            "data": img_base64,
                        },
                    },
                    {
                        "type": "text",
                        "text": "请详细描述这张图片的内容。",
                    },
                ],
            }
        ],
    },
    timeout=120,
)

result = response.json()
print(result["content"][0]["text"])
3.5 Python 示例(使用 Anthropic SDK)
import anthropic
import base64

client = anthropic.Anthropic(
    api_key="sk-你的密钥",
    base_url="https://www.moyu.info",
)

# 方式一:base64
with open("my_image.png", "rb") as f:
    img_base64 = base64.b64encode(f.read()).decode()

message = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "image",
                    "source": {
                        "type": "base64",
                        "media_type": "image/png",
                        "data": img_base64,
                    },
                },
                {
                    "type": "text",
                    "text": "请描述这张图片。",
                },
            ],
        }
    ],
)

print(message.content[0].text)

Logo

欢迎加入DeepSeek 技术社区。在这里,你可以找到志同道合的朋友,共同探索AI技术的奥秘。

更多推荐