# OCR

Optical Character Recognition (OCR) is the process of text recognition, used to extract text from images, serving as an easier alternative to manual data entry.

## Perform OCR

<mark style="color:blue;">`GET`</mark> `https://infer.roboflow.com/doctr/ocr`

#### Query Parameters

| Name                                       | Type   | Description                                                                                                    |
| ------------------------------------------ | ------ | -------------------------------------------------------------------------------------------------------------- |
| api\_key<mark style="color:red;">\*</mark> | String | [A Roboflow API Key](https://app.gitbook.com/s/e5GEiPeDoFksvZv1vH3A/authentication/find-your-roboflow-api-key) |

#### Request Body

| Name                                          | Type   | Description                                         |
| --------------------------------------------- | ------ | --------------------------------------------------- |
| image<mark style="color:red;">\*</mark>       | Object | A JSON object containing image information and data |
| image.type<mark style="color:red;">\*</mark>  | String | One of the following: "url" or "base64"             |
| image.value<mark style="color:red;">\*</mark> | String | Either a URL or a base64 encoded image              |

{% tabs %}
{% tab title="200: OK " %}

{% endtab %}
{% endtabs %}

To learn about how to deploy an OCR solution locally on your device using Python, [check out this blog post](https://blog.roboflow.com/ocr-api/).

{% hint style="success" %}
You can also deploy an OCR solution on your device using our open-source inference server. [Learn more about deploying DocTR (OCR) locally here.](https://inference.roboflow.com/foundation/doctr/)
{% endhint %}