Molmo2 Image Content Moderator API Documentation

Molmo2 Image Content Moderator API Documentation

Playground

Molmo2-4B Image Content Moderator: Analyze image content for safety, appropriateness, and policy compliance. Detects violence, nudity, gore, and other harmful visual content. Open-source vision-language model. Ready-to-use REST API, no cold starts, affordable pricing.

Features

Automatically screen images for harmful content with Molmo2 Image Content Moderator. This AI-powered moderation tool analyzes images and returns safety classifications for harassment, hate speech, sexual content, and violence — essential for content platforms, user-generated content, and compliance workflows.

Need video moderation? Try Molmo2 Video Content Moderator for video analysis.

Why It Works Great

• Comprehensive detection: Screens for harassment, hate, sexual content, and violence.
• Child safety: Dedicated detection for content involving minors.
• Instant results: Fast processing with results in seconds.
• JSON output: Clean, structured results for easy integration.
• Ultra-affordable: Just $0.003 per image — 333 images for $1.
• Custom criteria: Optional text input for additional context.

Parameters

How to Use

1. Upload your image — drag and drop or paste a public URL.
2. Add context (optional) — provide additional text for custom criteria.
3. Run — click the button to analyze.
4. Review results — check the JSON output for safety classifications.

Pricing

Flat rate per image analyzed.

Output Format

The model returns a JSON object with boolean flags for each content category:

{
 "harassment": false,
 "hate": false,
 "sexual": false,
 "sexual/minors": false,
 "violence": false
}

Detection Categories

Best Use Cases

• Content Platforms — Screen user-uploaded images before publishing.
• Social Media — Moderate image content at scale.
• E-commerce — Review product images for policy compliance.
• Dating Apps — Filter inappropriate profile photos.
• Forums & Communities — Ensure uploaded images meet guidelines.
• Pre-screening — Filter content before human review.

Pro Tips for Best Results

• Use as a first-pass filter before human moderation for edge cases.
• Integrate via API for automated moderation pipelines.
• At $0.003 per image, batch processing is extremely cost-effective.
• Combine with video moderation for comprehensive content screening.
• Use the text parameter to provide context for borderline content.

Notes

• If using a URL, ensure it is publicly accessible.
• Processing is near-instant for most images.
• Returns boolean values — true indicates detected content.
• Designed for automated workflows with JSON output.
• Consider human review for flagged content or edge cases.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/molmo2/image-content-moderator" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    created|processing) sleep 2 ;;
    *) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters