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

# Optical Character Recognition (OCR)

> Send images to our OCR engine and receive a low latency response of text in JSON format

<Tip>
  To create your free trial API key, you can head over to the [API Key Management Dashbord](https://dashboard.finvera.ai/sign-up).
</Tip>

### **Real-Time OCR**

* Ultra-Low Latency: Optimized for real-time use cases.
* Accurate OCR Models: Detects text, numbers, and symbols with confidence scoring.
* Structured Output: Returns clean JSON with text, coordinates, and metadata.

Example Request:

```json theme={null}
curl -X POST https://api.finvera.news/ocr/v1/extract \
-H "Content-Type: application/json" \
-d '{
"frame": "data:image/png;base64,..."
}'

```

Example Response:

```json theme={null}
{
"timestamp": "2024-11-05T17:00:00Z",
"results": [
{ "text": "Q3 Earnings Call", "x": 120, "y": 45, "confidence": 0.97 },
{ "text": "Revenue: $14.2M", "x": 128, "y": 78, "confidence": 0.94 }
]
}
```

### **Use Cases**

* Real-time video transcription
* Financial broadcast analysis
* Compliance and monitoring
* Visual data enrichment for AI pipelines

### **Video Stream Example**

When frames like the one below are sent to the endpoint, the OCR engine will extract the text and send via JSON.

<Frame>
  <img src="https://mintcdn.com/finvera/_NWQgeoPm_6y1fWM/assets/OCR.png?fit=max&auto=format&n=_NWQgeoPm_6y1fWM&q=85&s=2673e682c188d3286babaeb6bd030a61" alt="OCR on a video feed" width="1208" height="679" data-path="assets/OCR.png" />
</Frame>

Example Response:

```json theme={null}
{
"timestamp": "2024-11-05T17:36:02Z",
"results": [
{ "text": "BREAKING NEWS" },
{ "text": "BOEING DELAYS 777X ENTRY TO SERVICE" }
]
}
```