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

# Vision Recognition

> Multimodal interaction with visual recognition capabilities

<Frame>
  <img src="https://mintcdn.com/taamai-c8c27d6c/7tpU7wOp-rlopJuB/images/playground/vision-demo.png?fit=max&auto=format&n=7tpU7wOp-rlopJuB&q=85&s=be57b2fdbd8f667e562960e8ec755c1c" alt="Vision Recognition Demo" width="4320" height="2700" data-path="images/playground/vision-demo.png" />
</Frame>

## Overview

Gately AI supports advanced visual recognition capabilities through multiple state-of-the-art models, enabling rich multimodal interactions in your applications.

## Supported Models

<CardGroup cols={2}>
  <Card title="GPT-4 Vision" icon="eye">
    * High accuracy recognition
    * Detailed scene analysis
    * Complex visual reasoning
  </Card>

  <Card title="Gemini Pro Vision" icon="camera">
    * Fast processing
    * Multiple object detection
    * Context understanding
  </Card>
</CardGroup>

## Features

<AccordionGroup>
  <Accordion title="Image Upload" icon="upload">
    Multiple ways to input images:

    * Drag and drop
    * File upload
    * Image URL
    * Clipboard paste
  </Accordion>

  <Accordion title="Recognition Capabilities" icon="wand-magic-sparkles">
    What our models can analyze:

    * Object identification
    * Text in images (OCR)
    * Scene understanding
    * Visual relationships
    * Color and style analysis
  </Accordion>

  <Accordion title="Interaction Methods" icon="comments">
    Ways to interact with images:

    * Natural language queries
    * Visual question answering
    * Image description
    * Detail analysis
  </Accordion>
</AccordionGroup>

## Using Vision Recognition

<Steps>
  <Step title="Select Vision Model" icon="check">
    Choose from available vision-enabled models in the model selector
  </Step>

  <Step title="Upload Image" icon="upload">
    Drag and drop or click to upload your image

    <Frame>
      <img src="https://mintcdn.com/taamai-c8c27d6c/7tpU7wOp-rlopJuB/images/playground/upload-demo.png?fit=max&auto=format&n=7tpU7wOp-rlopJuB&q=85&s=a7be8a77e0b4add93ce935040856b314" alt="Upload Interface" width="4320" height="2700" data-path="images/playground/upload-demo.png" />
    </Frame>
  </Step>

  <Step title="Ask Questions" icon="message">
    Enter your queries about the image:

    * "What's in this image?"
    * "Can you describe the scene?"
    * "What text appears in this image?"
  </Step>
</Steps>

## Example Interactions

```json theme={null}
// Example API request
{
  "model": "gpt-4-vision",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "text",
          "text": "What's in this image?"
        },
        {
          "type": "image_url",
          "image_url": "https://example.com/image.jpg"
        }
      ]
    }
  ]
}
```

## Best Practices

<CardGroup cols={3}>
  <Card title="Image Quality" icon="image">
    Use clear, well-lit images
  </Card>

  <Card title="Specific Questions" icon="message-question">
    Ask focused, clear questions
  </Card>

  <Card title="Context Matters" icon="circle-nodes">
    Provide relevant context
  </Card>
</CardGroup>

<Note>
  Images are processed securely and temporarily. See our [privacy policy](/privacy) for details.
</Note>

<Warning>
  Vision recognition consumes additional credits based on image size and complexity.
</Warning>