Qufas Python SDK

Official Python SDK for managing Qufas AI Platform.

The Qufas Python SDK is a platform management SDK. It is designed for account and developer workflows such as listing models, reading usage, checking billing balance, and listing API key metadata.

For chat completions and model inference, use the official OpenAI SDK with the Qufas OpenAI-compatible endpoint:

from openai import OpenAI

client = OpenAI(
    api_key="qf_sk_xxxxx",
    base_url="https://api.qufas.ai/v1",
)

Installation

pip install qufas

Quick Start

from qufas import Qufas

client = Qufas(
    api_key="qf_sk_xxxxx"
)

client.models.list()
client.usage.list()
client.billing.balance()
client.api_keys.list()

You can also load the API key from an environment variable:

export QUFAS_API_KEY="qf_sk_xxxxx"

from qufas import Qufas

client = Qufas()
print(client.billing.balance())

Platform APIs

Models

List models available to the authenticated Qufas API key.

models = client.models.list()

for model in models["data"]:
    print(model["id"])

Usage

List recent usage records for the authenticated account.

usage = client.usage.list(limit=25)

print(usage["totals"]["total_tokens"])
print(usage["totals"]["cost"])

Billing

Read the current prepaid balance.

balance = client.billing.balance()

print(balance["balance"], balance["currency"])

API Keys

List API key metadata. Full API key values are never returned.

keys = client.api_keys.list()

for key in keys["data"]:
    print(key["id"], key["prefix"], key["status"])

client.api_keys.create() and client.api_keys.delete() are included for forward compatibility with the public API surface. They may return 501 not_implemented until Qufas enables public API key creation and deletion.

Chat Completions

Qufas AI is OpenAI-compatible. Use the official OpenAI SDK for model inference.

from openai import OpenAI

client = OpenAI(
    api_key="qf_sk_xxxxx",
    base_url="https://api.qufas.ai/v1",
)

response = client.chat.completions.create(
    model="qwen-plus",
    messages=[
        {
            "role": "user",
            "content": "Write a haiku about artificial intelligence.",
        }
    ],
)

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

The Qufas Python SDK intentionally does not duplicate chat completion methods. This keeps the primary developer experience compatible with existing OpenAI SDK applications.

Configuration

from qufas import Qufas

client = Qufas(
    api_key="qf_sk_xxxxx",
    base_url="https://api.qufas.ai",
    timeout=30.0,
)

Error Handling

from qufas import Qufas, QufasAPIError, QufasAuthenticationError

client = Qufas(api_key="qf_sk_xxxxx")

try:
    print(client.billing.balance())
except QufasAuthenticationError:
    print("Invalid or inactive API key.")
except QufasAPIError as error:
    print(error.status_code, error.error_code, str(error))

Exception classes:

• QufasError
• QufasAPIError
• QufasAuthenticationError
• QufasRateLimitError
• QufasServerError
• QufasConnectionError

Development

cd qufas-python
python -m pip install -e .
python -m unittest discover -s tests

License

MIT