Streaming Response API

Request

• Method: POST
• URL: /api/qwen-stream
• Headers:
  • Content-Type: application/json
• Body:

  {
    "imageBase64": "data:image/jpeg;base64,...",
    "prompt": "optional instruction, defaults to traffic accident detection"
  }

Streaming Response

• Content-Type: text/event-stream
• Data Format: Real-time text chunks returned from model generation
• End Marker: Usage information included at the end

  --- Usage ---
  {"prompt_tokens":10,"completion_tokens":20,"total_tokens":30}

Error Response

• Content-Type: application/json
• Format:

  {
    "error": "error message",
    "code": 400
  }

Error Codes

• 400: Request parameter error
• 500: Server internal error

Example

// Client handling streaming response
const response = await fetch('/api/qwen-stream', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    imageBase64: "data:image/jpeg;base64,...",
    prompt: "Analyze vehicle status in the image"
  })
});

const reader = response.body.getReader();
const decoder = new TextDecoder();

while (true) {
  const { done, value } = await reader.read();
  if (done) break;
  console.log(decoder.decode(value));
}

Notes

1. Images must be base64 encoded, supports data:image/jpeg;base64 prefix
2. Streaming responses require text/event-stream parsing
3. Default prompt is traffic accident detection
4. Response ends with token usage information

Code Examples

curl -X POST "http://segvision.satxspace.org/api/qwen-stream" \
-H "Content-Type: application/json" \
-d '{
  "imageBase64": "data:image/jpeg;base64,your-image-base64-data",
  "prompt": "Detect traffic accidents in the image and return results in JSON format."
}'