kestrel-sdk 0.1.0

Python SDK for Kestrel — route LLM requests to the cheapest capable model

Kestrel Python SDK

Drop-in LLM cost optimization. Route requests to the cheapest capable model with one line of code.

Install

pip install kestrel-sdk

Usage

import kestrel_sdk

client = kestrel_sdk.Client(
    api_key="ks-your-kestrel-key",
    provider_key="sk-your-openai-key",
    base_url="http://localhost:8080/v1",
)

# Send requests as usual — Kestrel handles routing
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "What is 2+2?"}],
)

The SDK is a thin wrapper over the OpenAI Python SDK. All openai.OpenAI methods work as expected.

Async

import kestrel_sdk

client = kestrel_sdk.AsyncClient(
    api_key="ks-your-key",
    provider_key="sk-your-openai-key",
)

response = await client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)

Learn More

• Kestrel documentation
• How routing works