Moderate Image

Analyze images for inappropriate visual content including nudity, violence, hate symbols, and more. Images are processed by Claude Vision (Tier 2) for high-accuracy classification.

Request

Provide the image as either a URL or base64-encoded data. Exactly one of url or data is required.

A publicly accessible URL pointing to the image. Supported formats: JPEG, PNG, GIF, WebP. Maximum file size: 10 MB.

Base64-encoded image data. Include the data URI prefix (e.g., data:image/png;base64,...). Maximum decoded size: 10 MB.

Optional list of categories to evaluate. Defaults to all available categories if omitted.

Supported values: sexual, violence, hate_symbols, self_harm, drugs, gore.

Response

Unique identifier for this moderation request.

Recommended action: allow, flag, or block.

true if any category score exceeded its threshold.

Per-category moderation results.

Confidence score from 0.0 to 1.0.

The configured threshold for this category.

true if the score exceeds the threshold.

Always "haiku" (Claude Vision) for image moderation.

Total processing time in milliseconds.

Examples

Moderate by URL

curl -X POST https://api.getsieve.dev/v1/moderate/image \
  -H "Authorization: Bearer mod_live_abc123..." \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/user-upload.jpg"
  }'

const response = await fetch("https://api.getsieve.dev/v1/moderate/image", {
  method: "POST",
  headers: {
    "Authorization": "Bearer mod_live_abc123...",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    url: "https://example.com/user-upload.jpg",
  }),
});

const result = await response.json();

import requests

response = requests.post(
    "https://api.getsieve.dev/v1/moderate/image",
    headers={"Authorization": "Bearer mod_live_abc123..."},
    json={"url": "https://example.com/user-upload.jpg"},
)

result = response.json()

Moderate by Base64

curl -X POST https://api.getsieve.dev/v1/moderate/image \
  -H "Authorization: Bearer mod_live_abc123..." \
  -H "Content-Type: application/json" \
  -d '{
    "data": "data:image/png;base64,iVBORw0KGgo..."
  }'

const response = await fetch("https://api.getsieve.dev/v1/moderate/image", {
  method: "POST",
  headers: {
    "Authorization": "Bearer mod_live_abc123...",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    data: "data:image/png;base64,iVBORw0KGgo...",
  }),
});

import requests

response = requests.post(
    "https://api.getsieve.dev/v1/moderate/image",
    headers={"Authorization": "Bearer mod_live_abc123..."},
    json={"data": "data:image/png;base64,iVBORw0KGgo..."},
)

Example Response

{
  "request_id": "req_img_9d8e7f",
  "action": "allow",
  "flagged": false,
  "categories": {
    "sexual": {
      "score": 0.03,
      "threshold": 0.8,
      "flagged": false
    },
    "violence": {
      "score": 0.01,
      "threshold": 0.7,
      "flagged": false
    },
    "hate_symbols": {
      "score": 0.0,
      "threshold": 0.7,
      "flagged": false
    },
    "self_harm": {
      "score": 0.0,
      "threshold": 0.7,
      "flagged": false
    },
    "drugs": {
      "score": 0.02,
      "threshold": 0.8,
      "flagged": false
    },
    "gore": {
      "score": 0.0,
      "threshold": 0.7,
      "flagged": false
    }
  },
  "pipeline_tier": "haiku",
  "latency_ms": 420
}