> For the complete documentation index, see [llms.txt](https://docs.docbits.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.docbits.com/administration-and-setup/settings/document-processing/classification-and-extraction/ai-model.md).

# AI Model

## Overview

The **AI Model** setting allows you to define which AI model is used by default for **field extraction** and **table extraction** during document processing.\
In this section, you can review the token cost for each model and see which model is currently assigned to each supplier.

## How to Access

1. Navigate to **Settings** → **Document Processing** → **Classification and Extraction**

   <figure><img src="/files/CQdBOnDb3XsxJQ0yhB49" alt=""><figcaption></figcaption></figure>
2. Scroll down to the **Table Extraction** section

   <figure><img src="/files/2I3FkEDeiaZ4CKs7NuXa" alt=""><figcaption></figcaption></figure>

## AI Model Options

DocBits offers three AI model options for field and table extraction. They differ in the balance between **extraction accuracy**, **processing speed**, and **token cost per document** — so you can match the option to the kind of documents you process. Hover over the info icon next to the setting to see the token cost for the currently selected option.

* **Full** – The most thorough option, with the highest extraction accuracy. Best for complex layouts, low-quality scans, or documents where precision matters most. As the most powerful option it is also the slowest, at **2 tokens per document**.
* **Fast** – A balanced option that combines strong accuracy with quicker processing at a lower cost. This is the recommended default for most everyday documents, at **1 token per document**.
* **Turbo** – The quickest and most economical option. Best suited to high volumes of simple, clean, well-structured documents where speed and low cost matter more than maximum accuracy, at **1 token per document**.

| Option    | Best for                                          | Accuracy | Speed   | Token cost   |
| --------- | ------------------------------------------------- | -------- | ------- | ------------ |
| **Full**  | Complex layouts, poor scans, high-precision needs | Highest  | Slowest | 2 / document |
| **Fast**  | Everyday documents (recommended default)          | High     | Fast    | 1 / document |
| **Turbo** | High volumes of simple, clean documents           | Good     | Fastest | 1 / document |

<figure><img src="/files/f09eQgWpQdUPp1igsgB7" alt=""><figcaption></figcaption></figure>

## AI Model Assignment Table

You can also configure supplier-specific **AI models** directly in the **Validation screen**, allowing you to fine-tune extraction accuracy for individual suppliers.\
\
For more information, please refer to the corresponding documentation [here](/end-user-and-partner-section/end-user-section/validation-screen/supplier-specific-ai-model-for-field-and-table-extraction.md).

The assignment table displays the AI model settings for each supplier and includes the following details:

* **Supplier ID** – The unique identifier of the supplier
* **AI Model** – The AI model currently assigned to the supplier
* **E-Text**: Indicates whether the E-Text feature is enabled
* **Action** – Contains the option to delete the entry

<figure><img src="/files/og1NX2Df0kKzNoGUBEWn" alt=""><figcaption></figcaption></figure>

### Delete Entry – Reset Supplier-Specific Settings

To reset a supplier’s AI model setting to the default:

1. Click the trashcan icon in the **Action** column next to the supplier entry.

   <figure><img src="/files/ONFTaqExT2eQjOVry79R" alt=""><figcaption></figcaption></figure>
2. A confirmation dialog will appear—confirm that you want to delete the entry.

   <figure><img src="/files/CxI2mQDHsjVgOBoTEyvu" alt=""><figcaption></figcaption></figure>

Once deleted, the supplier will revert to using the default **AI model** for **field extraction** and **table extraction**.


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.docbits.com/administration-and-setup/settings/document-processing/classification-and-extraction/ai-model.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.