Claude Vision

Analyze images using Anthropic Claude’s vision capabilities.

Overview

Sends an image (from binary data, a URL, or a previously uploaded Anthropic file) along with a text prompt to Anthropic’s Claude Messages API (POST /v1/messages) with image content blocks. Returns the model’s text analysis of the image. Supports Claude Opus, Sonnet, and Haiku models with base64-encoded binary images, direct URL references, or file_id references. The response text is stored in a configurable output field (default: “response”) along with model name, token usage, and stop reason.

Category: AI
Tool Name: claude_vision
Version: 1

Appearance: Icon: anthropic | Color: #d4a574

Node Type

Action (Binary) — handles file/binary data operations

Input / Output

Direction	Port(s)
Input	Input
Output	Output, Error

Credentials

This tool requires Anthropic credentials. See the Credentials Guide for setup instructions.

Parameters

Parameter	Type	Required	Default	Description
Model	options	No	Claude Sonnet	The Claude model to use for vision analysis. Always uses the latest version (auto-updated).
				Options: Claude Opus (most capable — complex visual analysis and reasoning), Claude Sonnet (balanced — strong visual analysis at lower cost and latency), Claude Haiku (fastest — simple image tasks and high-volume work). Each option tracks the current release of its tier, so the underlying model ID updates without any change to your node.
System Prompt	string	No	—	Optional system prompt to set the model’s behavior and context for image analysis. Supports expressions.
User Message	string	No	Describe this image in detail	The text prompt to send alongside the image. Describes what analysis to perform. Supports expressions.
Image Source	options	No	binary	Where to get the image — from binary data, a URL, or a previously-uploaded Anthropic file_id.
				Options: binary (read the image from binary data attached to the input item), url (provide an image URL for Claude to analyze), file (reference an image previously uploaded to Anthropic via Claude File Upload)
Binary Property Name	string	No	data	The name of the binary property containing the image data. Names are case-sensitive — see the upstream node’s Binary Data panel for the exact names to use. (shown when Image Source is binary)
Image URL	string	No	—	URL of the image to analyze. Falls back to item.imageUrl if empty. Supports expressions. (shown when Image Source is url)
Anthropic File ID	string	No	—	The file_id returned by a previous Claude File Upload node. Only used when Image Source is “file”. Falls back to item.fileId if empty. Supports {{ $json.fileId }} expressions from an upstream Claude File Upload node. Supports expressions. (shown when Image Source is file)
Options	collection	No	{}	Advanced generation and output settings.
— Max Tokens	number	No	4096	Maximum number of tokens to generate in the response.
— Temperature	number	No	—	Sampling temperature (0-1). Lower values are more deterministic. Leave unset to use the model default.
— Response Field Name	string	No	response	The output field name where the vision analysis text will be stored.
Include Input	boolean	No	false	Whether to include the original input item fields in the output alongside the response.
Max Concurrency	number	No	5	Maximum number of items to process concurrently. Lower values recommended for vision due to larger payloads.

Output Data

One image per input item, and one output item per input item. The analysis text lands on the field named by Response Field Name (response by default). Binary data on the input item is forwarded unchanged — including the image itself — so downstream nodes can still work with the file. With Include Input on, the original item fields are merged in alongside the result.

{
  "response": "The model's description or analysis of the image",
  "model": "claude-...",
  "usage": { "input_tokens": 1620, "output_tokens": 288 },
  "stopReason": "end_turn"
}

• model is the model that actually answered, as reported by Anthropic.
• usage is the token accounting Anthropic returned for the call — images consume input tokens in proportion to their resolution.
• stopReason says why generation stopped — for example end_turn (finished naturally) or max_tokens (hit the Max Tokens cap).

Reference the result downstream by expression, e.g. {{ $json.response }}.

Usage Examples

• Describe what is in an uploaded image
• Extract text from a screenshot using Claude vision
• Classify product images by category
• Analyze charts and graphs from binary data
• Identify objects in photos from URLs

Example Configuration

Describe an image attached to the item as binary data:

{
  "type": "claude_vision",
  "parameters": {
    "imageSource": "binary",
    "binaryPropertyName": "data",
    "userMessage": "Describe this image in detail"
  }
}

Read structured data out of a chart supplied by URL:

{
  "type": "claude_vision",
  "parameters": {
    "imageSource": "url",
    "imageUrl": "{{ $json.chartUrl }}",
    "userMessage": "Extract all data points and labels from this chart.",
    "options": {
      "maxTokens": 2048,
      "temperature": 0,
      "responseFieldName": "chartData"
    }
  }
}

Document OCR from a scan on a custom binary property:

{
  "type": "claude_vision",
  "parameters": {
    "imageSource": "binary",
    "binaryPropertyName": "scan",
    "systemPrompt": "You are an expert OCR system. Extract text exactly as it appears, preserving formatting.",
    "userMessage": "Extract all text from this document image.",
    "includeInput": true,
    "options": {
      "maxTokens": 4096,
      "temperature": 0,
      "responseFieldName": "extractedText"
    }
  }
}

Analyze an image already uploaded by a Claude File Upload node:

{
  "type": "claude_vision",
  "parameters": {
    "imageSource": "file",
    "imageFileId": "{{ $json.fileId }}",
    "userMessage": "Identify any visible defects in this product image.",
    "options": {
      "maxTokens": 512,
      "responseFieldName": "defectAnalysis"
    }
  }
}

Error Handling

Mode	Behavior
stop	Halts workflow on first error
continue	Skips failed items, passes successful ones through
errorPort	Routes failed items to Error output port

Tips

Claude Vision analyzes images by sending binary data, a URL reference, or an Anthropic file_id from an upstream Claude File Upload node alongside a text prompt to Anthropic Claude via the Messages API. Use it when a workflow needs to extract information, describe content, or answer questions about visual inputs using the Opus, Sonnet or Haiku tier. It produces a response text field containing the model analysis, plus model name, token usage counts, and stop reason on the main output.