Streaming - QWave API

Overview

Text generation endpoints support Server-Sent Events (SSE) streaming. Set stream: true to receive output token by token as it’s generated.

Streaming-Enabled Endpoints

/v1/chat/completions — OpenAI compatible
/v1/messages — Claude Messages
/v1/responses — OpenAI Responses API

Example

The examples, choices, [DONE], and final-frame usage.cost below apply to /v1/chat/completions. Claude Messages and OpenAI Responses use different event structures; follow their endpoint documentation. Replace YOUR_MODEL with a text model available to your account.

from openai import OpenAI

client = OpenAI(
    base_url="https://api.qingbo.ai/v1",
    api_key="your-api-key"
)

stream = client.chat.completions.create(
    model="YOUR_MODEL",
    messages=[{"role": "user", "content": "写一首短诗"}],
    stream=True
)

for chunk in stream:
    if chunk.usage is not None:
        quota_cost = getattr(chunk.usage, "cost", None)
        if quota_cost is not None:
            print(f"\nQuota cost: {quota_cost} quota")
    if chunk.choices:
        content = chunk.choices[0].delta.content
        if content:
            print(content, end="")

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://api.qingbo.ai/v1',
  apiKey: 'your-api-key'
});

const stream = await client.chat.completions.create({
  model: 'YOUR_MODEL',
  messages: [{ role: 'user', content: '写一首短诗' }],
  stream: true
});

for await (const chunk of stream) {
  if (chunk.usage?.cost != null) {
    process.stdout.write(`\nQuota cost: ${chunk.usage.cost} quota\n`);
  }
  const content = chunk.choices?.[0]?.delta?.content;
  if (content) process.stdout.write(content);
}

SSE Data Format

Each Chat Completions chunk is a data: {json} SSE event:

data: {"id":"chatcmpl-xxx","choices":[{"delta":{"content":"你"},"index":0}]}

data: {"id":"chatcmpl-xxx","choices":[{"delta":{"content":"好"},"index":0}]}

data: [DONE]

When the stream ends, you’ll receive data: [DONE].

Final Frame Usage & Billing (cost)

When usage output is enabled for Chat Completions, the final usage frame before [DONE] carries the call’s usage and billed quota. Its choices array may be empty, so check the array before accessing choices[0]:

data: {"id":"chatcmpl-xxx","choices":[],"usage":{"prompt_tokens":12,"completion_tokens":8,"total_tokens":20,"cost":150}}

data: [DONE]

usage.cost is the call’s billed quota (an integer), not US dollars (USD); 150 means 150 quota
It is sent by default; explicitly passing stream_options: {"include_usage": false} disables it
With the OpenAI SDK, the final chunk’s chunk.usage contains this data
If the connection drops or no usage frame arrives, do not treat the missing cost as zero; check the console usage records