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

# Tazo API

> Turn spoken audio into typed decisions, such as which tool to call, with calibrated probabilities.

The Tazo API takes a short audio clip and a set of questions, and returns typed
answers with probabilities. You decide the questions and the choices. Tazo
decides which choice the speaker meant, and tells you how sure it is.

> **[Quickstart](/quickstart.md)**
>
> Create an API key and make your first decision in a few minutes.

> **[Authentication](/authentication.md)**
>
> Sign-in tokens, API keys, and which one each endpoint accepts.

> **[Decisions](/guides/decisions.md)**
>
> Question types, request shape, and how to read answers.

> **[API reference](/api-reference/introduction.md)**
>
> Every endpoint, with an interactive playground.

## How it works

```text theme={null}
audio + state + questions  ->  Tazo  ->  answers + probabilities
```

A request has three parts:

* **Audio**: one audio file, up to 16 MB.
* **State**: shared context for the decision, as a string, object, or list of strings.
* **Questions**: one or more questions with IDs you choose. Tazo returns each
  answer under the same ID.

```json Example answer theme={null}
{
  "action": {
    "type": "choice",
    "choice": "get_weather",
    "probabilities": { "NO_TOOL": 0.08, "get_weather": 0.92 },
    "confidence": 0.92
  }
}
```

## Base URL

```text theme={null}
https://api.tazolabs.com
```

All endpoints are under `/v1`, except `/healthz`. Fields may be added within v1,
but existing fields and answer types are never renamed or removed. A breaking
change ships as a new versioned endpoint.


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