> For the complete documentation index, see [llms.txt](https://langtech-bsc.gitbook.io/applied-intelligence/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://langtech-bsc.gitbook.io/applied-intelligence/models/decision-models.md).

# Decision models

Models that read a text and return a calibrated decision (a category, a yes/no, a score) instead of generating text.

Many tasks people solve with a chat LLM are really **decisions**:

* Which department should handle this message?
* Is this request harmful?
* How urgent is it?
* Should the assistant call a tool or ask the user first?

Asking a general LLM to *write* the answer is slow, hard to parse and impossible to trust: it never tells you how sure it is.

A **decision model** is built for this. You give it a text (or a JSON state) and a set of typed questions. It returns, for each question, the chosen option together with a **calibrated confidence**. It reads its answer from the probabilities of the options, so no text is generated, and that makes it fast.

## Recommended models

| Model                                                                                                         | Size | Language                                | Question types          | Status                                             |
| ------------------------------------------------------------------------------------------------------------- | ---- | --------------------------------------- | ----------------------- | -------------------------------------------------- |
| [**CatDecide-1**](https://langtech-bsc.gitbook.io/applied-intelligence/models/decision-models/catdecide-1.md) | 9B   | Catalan first, also Spanish and English | Choice · Yes/no · Score | <mark style="color:orange;">**Coming soon**</mark> |

## When to use a decision model instead of an LLM

| Use a decision model when…                                                         | Use an LLM when…               |
| ---------------------------------------------------------------------------------- | ------------------------------ |
| The answer is one of a known set of options                                        | The answer is free text        |
| You need a confidence score to decide what to automate and what to send to a human | A best-effort answer is enough |
| You ask many questions about the same text                                         | You ask one open question      |
| Latency and cost matter (classification, routing, moderation at scale)             | You need reasoning written out |

{% hint style="info" %}
**A natural pairing:** put a decision model in front of an [LLM](https://langtech-bsc.gitbook.io/applied-intelligence/models/text-models.md) to route requests, screen harmful inputs and replies, and decide whether a tool call is needed. Let the LLM do only the generating.
{% endhint %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://langtech-bsc.gitbook.io/applied-intelligence/models/decision-models.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
