OpenAI Moderation

Classify text for harmful content using OpenAI’s moderation models.

Overview

OpenAI Moderation calls the OpenAI Moderations API (POST /moderations) to classify text for harmful content categories such as sexual, hate, harassment, self-harm, violence, and more. Returns flagged status, per-category boolean flags, and per-category confidence scores. Model selection is configurable. Each input item is classified independently. Results are placed in a configurable output field (default: “moderation”) along with a top-level “flagged” boolean.

Category: AI
Tool Name: openai_moderation
Version: 1

Appearance: Icon: openai | Color: #10a37f

Node Type

Action — processes input items and produces output

Input / Output

Direction	Port(s)
Input	Input
Output	Output, Error

Credentials

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

Parameters

Parameter	Type	Required	Default	Description
Model	options	No	(current default)	The moderation model to use for classification. Always uses the latest version (auto-updated).
				Options: the OpenAI moderation models available to your workspace — pick one from the dropdown.
Text	string	Yes	—	The text to classify for harmful content. Falls back to item.json.text or item.json.content if empty. Supports expressions.
Options	collection	No	{}	Optional output settings — add only the fields you need.
— Response Field Name	string	No	moderation	Field name in the output JSON where the moderation results will be placed.
Include Input	boolean	No	false	Whether to include the original input item fields in the output alongside the moderation results.
Max Concurrency	number	No	10	Maximum number of items to process concurrently.

Output Data

One output item per input item. The full result lands on the field named by Response Field Name (moderation by default), and the overall verdict is repeated at the top level as flagged so you can branch on it directly. The rest of the input item JSON is dropped unless Include Input is on; binary data on the input item is forwarded unchanged.

{
  "moderation": {
    "flagged": false,
    "categories": { "hate": false, "violence": false },
    "categoryScores": { "hate": 0.0002, "violence": 0.0011 }
  },
  "flagged": false
}

• categories holds one boolean per harm category the model evaluated; categoryScores holds the matching confidence score for each. The exact category names come from the moderation model.
• flagged is true when any category tripped. Wire it into a Switch or Filter node to gate the rest of the pipeline.

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

Usage Examples

• Check user-generated content for policy violations
• Filter incoming messages for harmful content before processing
• Classify text as safe or flagged with per-category scores
• Screen customer reviews for hate speech or harassment
• Build a content moderation pipeline for user submissions

Example Configuration

Screen each item’s text with the defaults:

{
  "type": "openai_moderation",
  "parameters": {
    "text": "{{ $json.text }}"
  }
}

Moderate at volume and keep the original record for routing:

{
  "type": "openai_moderation",
  "parameters": {
    "text": "{{ $json.comment }}",
    "includeInput": true,
    "maxConcurrency": 25,
    "options": {
      "responseFieldName": "safety"
    }
  }
}

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

OpenAI Moderation classifies text against harmful content categories including sexual, hate, harassment, self-harm, and violence by calling the OpenAI Moderations API. Use it when a workflow needs to screen user-generated content before storing, routing, or displaying it to enforce content safety policies. It outputs a top-level flagged boolean along with a configurable moderation field containing per-category boolean flags and confidence scores for every evaluated harm category.