OPENAI规格 API接口使用说明
基本信息
- 模型名称: Qwen2.5-VL-32B-Instruct (多模态视觉语言模型)
- API URL:
https://openai/v1/chat/completions - 请求方法: POST
认证方式
Bearer Token
在请求头 (Header) 中加入 Authorization 字段进行认证。
Authorization: Bearer your_api_key_here
请求示例
1. 纯文本对话
curl -X POST "https://openai/v1/chat/completions" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your_api_key_here" \
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "user",
"content": "你好"
}
]
}'
2. 图像理解 (Base64编码)```bash
curl -X POST "https://openai/v1/chat/completions"
-H "Content-Type: application/json"
-H "Authorization: Bearer your_api_key_here"
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "user",
"content": [
{
"type": "text",
"text": "请描述这张图片的内容"
},
{
"type": "image_url",
"image_url": {
"url": "image/path"
}
}
]
}
]
}'
### 3. 图像理解 (URL方式)
```bash
curl -X POST "https://openai/v1/chat/completions" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your_api_key_here" \
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "user",
"content": [
{
"type": "text",
"text": "这张图片展示了什么?"
},
{
"type": "image_url",
"image_url": {
"url": "https://example.com/image.jpg"
}
}
]
}
]
}'```
### 4. 多轮对话
```bash
curl -X POST "https://openai/v1/chat/completions" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your_api_key_here" \
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "system",
"content": "你是一个专业的图像分析助手"
},
{
"role": "user",
"content": "你好"
},
{
"role": "assistant",
"content": "你好! 我是一个专业的图像分析助手,可以帮你分析和理解图片内容。"
},
{
"role": "user",
"content": "请帮我分析这张图片中的物体"
}
]
}'
5. 流式响应
通过将 stream 参数设置为 true 来实现流式响应。
curl -X POST "https://openai/v1/chat/completions" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your_api_key_here" \
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "user",
"content": "请写一个关于人工智能的短文"
}
],
"stream": true
}' \
--no-buffer
6. 图像+文本复合分析
curl -X POST "https://openai/v1/chat/completions" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your_api_key_here" \
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "user",
"content": [
{
"type": "text",
"text": "请分析这张图片中的技术架构,并提供改进建议:"
},
{
"type": "image_url",
"image_url": {
"url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAA..."
}
}
]
}
]
}'
7. 使用API Key参数认证
除了 Bearer Token，也支持通过 URL 参数传递 API Key。
curl -X POST "https://openai/v1/chat/completions?apikey=your_api_key_here" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen2.5-v1-32b-instruct",
"messages": [
{
"role": "user",
"content": "你好,请介绍一下你的功能"
}
]
}'
8. 本地图片转Base64示例
# 首先将本地图片转换为Base64
IMAGE_BASE64=$(base64 -i /path/to/your/image.jpg)
# 然后发送请求
curl -X POST "https://openai/v1/chat/completions" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer your_api_key_here" \
-d "{
\"model\": \"qwen2.5-v1-32b-instruct\",
\"messages\": [
{
\"role\": \"user\",
\"content\": [
{
\"type\": \"text\",
\"text\": \"请分析这张图片\"
},
{
\"type\": \"image_url\",
\"image_url\": {
\"url\": \"data:image/jpeg;base64,${IMAGE_BASE64}\"
}
}
]
}
]
}"
请求参数说明
必需参数
model(string): 模型名称, 固定为"qwen2.5-v1-32b-instruct"。messages(array): 对话消息数组, 详见Messages格式说明。
可选参数
temperature(float): 控制随机性，值越高随机性越强。max_tokens(integer): 最大生成token数。stream(boolean): 是否流式返回。top_p(float): 核采样参数。frequency_penalty(float): 频率惩罚。presence_penalty(float): 存在惩罚。stop(string|array): 停止词。
Messages 格式
纯文本消息
{
"role": "user",
"content": "你的问题"
}
多模态消息 (文本+图像)
{
"role": "user",
"content": [
{
"type": "text",
"text": "请分析这张图片"
},
{
"type": "image_url",
"image_url": {
"url": "data:image/jpeg;base64,base64_encoded_image"
}
}
]
}
支持的图像格式
- JPEG
- PNG
- WebP
- GIF
响应格式
标准响应
{
"id": "chatcmpl-xxx",
"object": "chat.completion",
"created": 1234567890,
"model": "qwen2.5-v1-32b-instruct",
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": "模型的回复内容"
},
"finish_reason": "stop"
}
],
"usage": {
"prompt_tokens": 10,
"completion_tokens": 20,
"total_tokens": 30
}
}
流式响应
当 stream: true 时, 响应为 Server-Sent Events 格式:
data: {"id":"chatcmpl-xxx","object":"chat.completion.chunk","created":1234567890,"model":"qwen25-v1-32b-instruct","choices":[{"index":0,"delta":{"content":"你"},"finish_reason":null}]}
data: {"id":"chatcmpl-xxx","object":"chat.completion.chunk","created":1234567890,"model":"qwen25-v1-32b-instruct","choices":[{"index":0,"delta":{"content":"好"},"finish_reason":null}]}
data: [DONE]
错误处理
常见错误码
400: 请求参数错误401: 认证失败429: 请求频率超限500: 服务器内部错误502/503/504: 服务暂时不可用
错误响应格式
{
"error": {
"message": "错误描述",
"type": "invalid_request_error",
"code": "invalid_api_key"
}
}