> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tacetai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Make your first decision in about five minutes.

Tacet runs on our servers, so there’s nothing to download and no GPU to manage.

<Steps>
  <Step title="Get an API key">
    Sign in to the [Tacet console](https://console.tacetai.com), open API keys, and create one. Store it as an environment variable.

    ```bash theme={null}
    export TACET_API_KEY="tk_live_..."
    ```
  </Step>

  <Step title="Install the SDK">
    ```bash theme={null}
    pip install tacet
    ```
  </Step>

  <Step title="Ask a question">
    Describe the situation in `state`, then list the questions you want answered.

    ```python theme={null}
    from tacet import Tacet

    client = Tacet()  # reads TACET_API_KEY

    result = client.decide(
        model="tacet-alpha",
        state="My card was declined at checkout, but I can still log in.",
        questions={
            "route": {
                "type": "choice",
                "instructions": "Which team should handle this?",
                "criteria": {
                    "payments": "Card and billing problems",
                    "account": "Login and access",
                },
            }
        },
    )

    route = result.answers["route"]
    print(route.top, route.confidence)  # payments 0.97
    print(route.probs)                  # {'payments': 0.97, 'account': 0.03}
    ```
  </Step>

  <Step title="Act on it">
    Use the confidence to decide whether to act automatically or hand off. See [Thresholds and calibration](/guides/calibration).
  </Step>
</Steps>

<Note>
  Every answer is one of the options you listed, with a probability for each. The model never writes free text, so there’s nothing to parse.
</Note>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.