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

# Text Recognition

> Add professional OCR to your product: extract text and structured fields from images.

## What it is

**Text Recognition** is a new skill in the Skill Library: it reads text from images and, where possible, returns usable structured fields.

Think of it this way: users upload an ID, invoice, or product photo in your app; the system recognizes the text and fills forms or continues the flow—no separate third-party OCR integration and no manual typing.

**One-line positioning:** Integrate professional text recognition into your product—extract text and structured information from images with precision.

***

## Good for

Common scenarios:

* **Document review**: ID cards, passports, driver licenses, vehicle licenses, business licenses, and more
* **Finance / receipts**: VAT invoices, mixed receipts, bank cards, and more
* **Operations**: license plates, tables, text on product images
* **General extraction**: printed and handwritten text

Better than “just ask an AI to look at the image” for documents: IDs and invoices aim for field-level results (name, number, amount, etc.)—more accurate and usually more efficient.

***

## Supported types (\~12)

Categories called out in the product include:

* General text
* Handwritten text
* ID card
* Bank card
* Vehicle license
* Driver license
* Passport
* Business license
* VAT invoice
* Mixed receipts
* License plate
* Table

Covers high-frequency ecommerce, operations, and review needs.

***

## Core capabilities

| Capability                 | Meaning                                                                          |
| -------------------------- | -------------------------------------------------------------------------------- |
| Multi-category recognition | Not only “there is text,” but specialized recognition by document / invoice type |
| Structured extraction      | IDs and invoices return fields when possible for auto-fill                       |
| Per-call billing           | Charged only on success; failures are not charged                                |
| Billable & traceable       | Usage maps to a dedicated OCR line item in Compute Credit billing                |
| Privacy-conscious          | Platform does not retain original images or recognition results                  |

***

## Where to find it

1. Open any project.
2. Switch to the **Build** tab.
3. Open **Skill Library** in the left nav.
4. Find **Text Recognition**.

***

## How to enable (3 steps)

1. Say in chat that you need ID / invoice recognition, or open the skill page and click **Use** (top right).
2. superun connects OCR to the project for you (official skill, fast enablement).
3. Let users upload images in your product to run recognition.

After enablement, the skill detail has **Intro / Use** pages. The Use page shows enabled status and a persistent privacy notice.

***

## Billing

* **Per successful call**: charged once on success; failures are free.
* Cost comes from project / account Compute Credit usage; bills map to a **Text Recognition** category.
* Unit price follows the live bill (not hard-coded in the UI).

***

## Privacy & compliance (required)

In-product notices include:

* IDs and invoices contain sensitive personal data.
* The platform does not retain original images or recognition results.
* Before using someone else’s ID image, confirm you have their authorization.

For KYC, account opening, or expense products, add authorization copy in your own product as well.

***

## vs. integrating a cloud OCR yourself

|         | superun Text Recognition                         | Self-integrated cloud OCR                 |
| ------- | ------------------------------------------------ | ----------------------------------------- |
| Setup   | Chat or click **Use**; auto-added to the project | Apply for keys, write APIs, configure env |
| Results | Multi-category + structured fields               | Depends on vendor and your wrapper        |
| Billing | Platform per-call; fail = free                   | Vendor bill                               |
| Privacy | Platform “no retention” commitment               | Depends on how you store data             |

Best for teams that want photo-to-text / ID recognition quickly without wiring a separate cloud OCR.

***

## FAQ

<AccordionGroup>
  <Accordion title="Am I charged when recognition fails?">
    No. Billing is per successful call only.
  </Accordion>

  <Accordion title="Where do I see this on the bill?">
    In Compute Credit billing, filter by consumption type for the Text Recognition mapping (as shown in product).
  </Accordion>

  <Accordion title="Does the platform keep my uploaded ID images?">
    No. Original images and results are not retained. You still need authorization before using someone else’s ID.
  </Accordion>
</AccordionGroup>