Use Qwen3.8-27B with the OpenAI Python SDK

Python · OpenAI-compatible API · Qwen3.8-27B

Before you start

You need a Prefill Systems API key. Existing pilot users can create one in the dashboard; other developers can request direct API access.

1. Install the client

python -m pip install openai

2. Configure the client

Keep the API key in an environment variable rather than source code.

# PowerShell
$env:PREFILL_API_KEY = "pf_your_key_here"

# bash
export PREFILL_API_KEY="pf_your_key_here"

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://prefillsystems.com/v1",
    api_key=os.environ["PREFILL_API_KEY"],
)

3. Create a completion

response = client.chat.completions.create(
    model="qwen/qwen3.8-27b",
    messages=[
        {"role": "system", "content": "Answer clearly and concisely."},
        {"role": "user", "content": "Why does prefix caching reduce prefill work?"},
    ],
)

print(response.choices[0].message.content)
print(response.usage)

The response includes prompt and completion token counts. Cached prompt tokens are reported separately when the inference worker identifies a cache hit.

4. Stream tokens

stream = client.chat.completions.create(
    model="qwen/qwen3.8-27b",
    messages=[{"role": "user", "content": "Write a short explanation of speculative decoding."}],
    stream=True,
    stream_options={"include_usage": True},
)

for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)
    if chunk.usage:
        print("\nusage:", chunk.usage)

Streaming uses server-sent events. A final usage-bearing chunk arrives before the [DONE] marker.

Compatibility notes

The API base is https://prefillsystems.com/v1.
The model ID is qwen/qwen3.8-27b.
Capacity pressure returns HTTP 429 rather than holding requests in a long queue.
Prompts and completions are not persisted; account and usage metadata is retained separately.

Next: see the Qwen3.8 tool-calling guide or the complete API documentation.