One API key.
One balance.
Every model.

ModelBeat routes each request to the model that should serve it, then hands back the response with the exact cost attached. You send auto; the router decides. It speaks the OpenAI API, so most integrations are a two-line change.

Quickstart API reference

Two metered endpoints, streaming supported. Read the limits first.

pythonquickstart

1from openai import OpenAI
2
3client = OpenAI(
4    api_key=os.environ["MODELBEAT_API_KEY"],
5    base_url="https://api.modelbeat.ai/v1",
6)
7
8# "auto" or a tier — you never name a model.
9r = client.chat.completions.create(
10    model="auto",
11    messages=[{"role": "user", "content": "hello"}],
12)
13
14# what the call cost you, in USD
15print(r.usage.cost["total_cost"])

What comes back with every call

model: autoOne routing interfaceNo model to pick, pin, or keep up with. The cheapest candidate that clears your request's hard gates serves it.Read →usage.costWhat it costPer-request USD, broken down by component. Reserved before routing, settled after.Read →X-Request-IdA thread to pullOn success and on failure. Quote it and we can find the exact call.Read →

One API key.One balance.Every model.

One API key.
One balance.
Every model.