Skip to content
Visionati Visionati Docs

Getting Started

Before using the API, you’ll need an account, credits, and an API key. See Getting Started if you haven’t set that up yet.

Your First Request

Section titled “Your First Request”
Terminal window
curl -X POST "https://api.visionati.com/api/fetch" \
  -H "X-API-Key: Token YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com/photo.jpg"}'

This sends the image to all your default backends with all features enabled and returns a JSON response with descriptions, tags, NSFW scores, and more.

Narrow Your Request

Section titled “Narrow Your Request”

You don’t always need everything. Specify only the backends and features you want to reduce cost and response time:

Terminal window
curl -X POST "https://api.visionati.com/api/fetch" \
  -H "X-API-Key: Token YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/photo.jpg",
    "backend": ["openai", "googlevision"],
    "feature": ["descriptions", "tags"]
  }'

Use a Role

Section titled “Use a Role”

Roles control how LLM backends describe your image. The default is general, but you can use specialized roles:

Terminal window
curl -X POST "https://api.visionati.com/api/fetch" \
  -H "X-API-Key: Token YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com/product.jpg", "role": "ecommerce"}'

See Roles for all options.

Use a Custom Prompt

Section titled “Use a Custom Prompt”

For full control, pass your own prompt. This overrides the role:

Terminal window
curl -X POST "https://api.visionati.com/api/fetch" \
  -H "X-API-Key: Token YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com/photo.jpg", "prompt": "List every plant species visible"}'

What’s Next

Section titled “What’s Next”