Transforms a_vision

Vision Analysis

Analyses an image (URL/base64) against a natural-language prompt.

What does it do?

Analyzes images using visual AI (e.g. GPT-4 Vision).

How does it work?

You pass in an image and a question (e.g. "What color is this shirt?"). The AI 'sees' the photo and answers you.

Why use it?

Super useful when a customer sends a photo of a product they want to buy and says: "Do you have this same model?"

Inputs

name	type
`image_url`	string
`prompt`	string

Outputs

name	type
`text`	string
`provider`	string
`model`	string
`usage`	object

Branches

Fixed branches

out — Output

Configuration

model string required default: "google/gemini-2.5-flash-lite"

image_url_field string required default: "image_url"

image_base64_field string required default: "image_base64"

image_urls_field string optional

prompt_field string required default: "prompt"

system_prompt string optional

max_tokens integer required default: 500

temperature number required default: 0.3

memory_key string optional

Default configuration

{
  "model": "google/gemini-2.5-flash-lite",
  "image_url_field": "image_url",
  "image_base64_field": "image_base64",
  "prompt_field": "prompt",
  "max_tokens": 500,
  "temperature": 0.3
}

Emitted events

This node does not publish events on the pubsub channel.