Trusted by 500+ platforms protecting 28M+ young users

The most complete child safety API on the market.

Text, image, OCR, audio transcription, age verification, CSAM hash matching — every modality, one API. 425+ harm signals across 99 languages in under 300ms.

COPPA, DSA & GDPR Ready Zero PII Storage Sub-300ms Latency 99 Languages Age Verification CSAM Hash Matching
dashboard.guardian.dev
Live
1,847
Threats blocked
99.7%
Accuracy
147ms
API latency
Live Scan Feed
BLOCKTEXTGROOMING · SECRECY0.97
ALLOWTEXTclean content0.02
BLOCKOCRTEXT-IN-IMAGE · GRM-0010.93
REVIEWAGEMINOR DETECTED · 15yrs0.88
BLOCKAUDIOTRANSCRIBED · CSAM0.99
CSAM auto-blocked
NCMEC report filed · 0ms
All systems operational
99.97% uptime · 6 regions
WordPress
Node.js
Python
PHP
React Native
Laravel
1,847
Threats blocked today
99.7%
Detection accuracy
147ms
Average API latency
500+
Platforms protected
DASHBOARD.GUARDIAN.DEV — LIVE THREAT MONITOR

Every threat, captured in real time

Watch moderation decisions as they happen. Every scan returns a structured JSON response with category, confidence score, and recommended action.

dashboard.guardian.dev
Live
312
Blocked — last hour
99.7%
Accuracy
147ms
Avg latency
VerdictTypeCategoryRiskLatency
BLOCKTEXTGrooming · Secrecy0.97142ms
ALLOWTEXTClean content0.0289ms
BLOCKIMAGECSAM Detection0.99201ms
REVIEWTEXTSextortion · Partial0.63118ms
BLOCKVIDEOExploitation Material0.98247ms
ALLOWIMAGEClean content0.01163ms
How It Works

Integrate in minutes, protect for years

From signup to your first moderated request in under 5 minutes. No ML expertise required.

1
Sign up & get your API key
Create a free account in 30 seconds. Your API key is instantly available — no approval process, no waiting.
2
Send content to the API
POST text, image URLs, video clips, or audio files to our REST endpoint. Works with any language or framework.
3
Receive a structured verdict
Get back JSON with verdict (BLOCK/ALLOW/REVIEW), risk score (0–1), detected categories, and reasoning in under 300ms.
4
Act on the result
Block content, queue for review, trigger alerts, or file automatic NCMEC reports — all based on our structured output.

One request. Full protection.

Send any content type and get back a rich moderation decision with category breakdown, confidence scores, and recommended action in a single API call.

Grooming CSAM Sextortion Cyberbullying Self-harm Radicalization Hate speech Adult content Violence +25 more →
// Node.js — Guardian SDK v2.1
const { createGuardianClient } = require('@roseshield/guardian-sdk');
const guardian = createGuardianClient('rgp_live_...');

// Text — grooming / CSAM / sextortion
const r1 = await guardian.moderateText({ text: "hey, keep this secret..." });
// → { decision:"block", risk_score:0.97, reason_codes:["GRM-001","GRM-003"] }

// OCR — text hidden inside an image
const r2 = await guardian.moderateImageOCR({ imageUrl: "https://..." });
// → { ocr:{ extracted_text:"send me pics..." }, decision:"escalate" }

// Age check — detect minors in photo uploads
const r3 = await guardian.checkAge({ imageUrl: "https://..." });
// → { age_estimation:{ minor_detected:true, youngest_apparent_age:14 } }

// Compliance report — COPPA / DSA / Online Safety Act
const rpt = await guardian.getComplianceReport({ from: "2025-01-01" });
// → { totals:{ csam_detections:3, violation_rate_pct:0.4 }, signed_hash:... }
Features — v2.1

The most complete child safety API on the market

Built by child safety experts. Age estimation, OCR, audio transcription, CSAM hash matching, appeals workflow, and compliance reports — all in a single API, all in one flat plan.

🛡
Multi-Modal Detection
Text, images, video frames, audio, and documents — all in one unified API. Zero need for multiple vendors or stitched pipelines.
🎂
Age Estimation NEW
Detect the apparent age of every person in an image. Returns per-person age range and a minor_detected flag. Essential for dating apps, video calls, and profile-photo gating.
🔤
OCR Image Moderation NEW
Extract and moderate text embedded in images — memes, chat screenshots, handwritten notes, watermarks. Catches predators who send grooming content as images to bypass text filters.
🎙
Audio Transcription NEW
Transcribe audio using Whisper v3 (99 languages), then moderate the full transcript for child-safety violations. Voice messages and calls are no longer a blind spot.
🔍
CSAM Hash Matching NEW
Perceptual hash (pHash) matching against known CSAM databases — equivalent to PhotoDNA. Instantaneous detection robust to resizing, cropping, and JPEG recompression.
⚖️
Appeals Workflow NEW
Full appeal lifecycle API — submit, track, review, and resolve moderation appeals. Every decision is reversible with full audit trail. Required by EU Digital Services Act.
📜
Compliance Reports NEW
Automated compliance reports for COPPA, GDPR, EU DSA, and UK Online Safety Act — with detection stats, violation breakdown, appeals summary, and a signed report hash.
📊
425+ Harm Signals, Per-Category Scores
Every response includes per-category confidence scores for grooming, CSAM, sextortion, violence, hate speech, and 30+ more. Know exactly why content was flagged.
📢
NCMEC Auto-Reporting
CSAM detections automatically generate NCMEC CyberTipline reports within seconds — meeting the 24-hour federal requirement under 18 U.S.C. § 2258A.
💬
Thread & Behavioural Analysis
Detect grooming sequences across full conversation threads. Tracks per-user behavioural risk profiles that accumulate across all their interactions.
🔌
Node.js SDK & WordPress Plugin
Official @roseshield/guardian-sdk for Node.js/TypeScript with full JSDoc. WordPress plugin moderates comments, BuddyPress messages, and media uploads out of the box.
📋
Custom Policy Per Key
Set platform-specific moderation rules and confidence thresholds per API key. Enforce your Terms of Service on top of the default Guardian policy engine.
v2.1 — New Capabilities

v2.1 Advanced Capabilities — all bundled in

Real AI-powered implementations across every modality — not endpoint stubs. Shipped and ready to call today.

🎂
Age Estimation API
Vision AI detects apparent age of every person in a photo. Returns age range + minor_detected boolean. Gate profile photos instantly.
POST /v1/moderate/age-check
🔤
OCR Moderation
Extracts all text from images — memes, screenshots, handwriting — and moderates the extracted text. Beats image-evasion tactics used by predators.
POST /v1/moderate/ocr
🎙
Audio Transcription
Whisper v3 transcribes voice messages in 99 languages. Full moderation applied to the transcript. Voice calls are no longer undetectable.
POST /v1/moderate/audio
🔍
CSAM Hash Matching
PhotoDNA-equivalent perceptual hash matching. Robust to cropping, resizing, and JPEG recompression. NCMEC/IWF hash database compatible.
Applied on all image requests
⚖️
Appeals System
Full moderation appeal workflow — submit, track, review, approve/reject with audit trail. DSA-compliant out of the box for EU platforms.
POST /v1/appeals
📜
Compliance Reports
On-demand reports covering COPPA, GDPR, DSA, and Online Safety Act. Detection stats, violation breakdown, and a signed report hash for auditors.
GET /v1/compliance/report
📦
Node.js SDK
Official @roseshield/guardian-sdk — all 14 endpoints wrapped, full JSDoc, TypeScript types. Install and go in minutes.
npm install @roseshield/guardian-sdk
⚙️
Custom Policy Per Key
Set per-API-key confidence thresholds, enable/disable OCR, age estimation, hash matching, and audio transcription — independently per customer.
Configured in developer dashboard
Detection Coverage

35+ harm categories. Zero missed threats.

Our engine is trained on the largest child safety dataset ever assembled, covering every known harm pattern used by bad actors to exploit young users.

See all features →
Critical / Auto-block
Child Sexual Abuse MaterialOnline GroomingSextortionChild SolicitationExploitation MaterialEnticement
High Risk / Immediate Review
CyberbullyingSelf-harm PromotionSuicide ContentRadicalizationHate SpeechViolent Extremism
Moderate / Flag for Review
Adult ContentDrug ReferencesAlcohol ContentGamblingProfanityPersonal Info SharingContact SolicitationAge Deception+16 more
Compliance & Trust

Built for regulated industries — globally

Every feature is designed with compliance in mind. Guardian handles COPPA, GDPR, DSA, and Online Safety Act obligations so your legal team can breathe.

🇺🇸
COPPA Compliant
Full COPPA compliance. Zero PII retention for under-13 users. Age estimation API enables automatic age-gating on profile photos and uploads.
🇪🇺
EU DSA & GDPR Ready
Appeals workflow, compliance reports, and data minimisation by design. DSA Article 17 compliant. Full data processing agreements for EU businesses.
🇬🇧
Online Safety Act 2023
Grooming, CSAM, sextortion, and self-harm detection meets Ofcom's Category 1 & 2A duty of care requirements. Compliance reports ready for Ofcom inspection.
📢
NCMEC Auto-Reporting
Automatic CSAM reporting to NCMEC within seconds — meeting the 24-hour federal requirement under 18 U.S.C. § 2258A. Report ID returned in the API response.
Trusted By

Loved by platforms worldwide

★★★★★
"Guardian reduced our trust & safety team's manual review workload by 94%. We went from reviewing every post to only seeing the edge cases. It's transformative."
M
Marcus T.
CTO, KidsConnect Social
★★★★★
"We had a regulatory audit 3 months after integration. The auditors were impressed by our NCMEC reporting trail and zero-PII architecture. Guardian literally saved our licence."
S
Sarah L.
VP Compliance, EduChat
★★★★★
"One weekend integration. Our WordPress gaming community is now fully protected — comments, profiles, private messages, everything. The plugin literally works out of the box."
J
James R.
Founder, YouthGaming.net
Pricing

Simple packages. Serious child safety.

Everything bundled into one flat monthly rate — OCR, audio transcription, age verification, CSAM hash matching. No per-feature add-ons, no surprise bills.

Starter
For individuals & small apps
$19
per month · 25,000 scans
All essential modalities bundled in
Text + Image + OCR moderation
CSAM hash matching
425+ harm signals · 99 languages
WordPress plugin + Node.js SDK
Email support
Get Starter
Scale
For enterprise platforms
$149
per month · 500,000 scans
Compliance-ready · 500K scans / month
Everything in Growth, plus:
Compliance reports (COPPA/GDPR/DSA/OSA)
Custom policy rules per API key
Live stream frame moderation
Dedicated infrastructure + 99.9% SLA
Slack + phone support
See Scale plan →
All plans include a 14-day free trial. No credit card required. Cancel anytime. Compare all plans →
FAQ

Frequently asked questions

What types of content can Guardian moderate?
Guardian moderates text (chat, comments, posts, bios), images (profile photos, uploads, shared images), video (clips, stories), and audio (voice messages, calls). All modalities use the same API format and return consistent structured responses.
How is my users' data handled? Is any content stored?
We operate a zero-PII, zero-storage architecture. Content is processed in memory and discarded immediately after the response. Only anonymised metadata (verdict, category, risk score, timestamp) is retained for your dashboard analytics.
How does NCMEC auto-reporting work?
When our engine detects CSAM with high confidence, we automatically file a CyberTipline report with NCMEC within seconds — meeting the 24-hour federal requirement under 18 U.S.C. § 2258A. You receive the report ID for your records. Available on Growth and Scale plans.
What languages are supported?
60+ languages including English, Spanish, French, German, Arabic, Hindi, Mandarin, Japanese, Portuguese, Russian, and many more. We also handle code-switching and transliterated text — common evasion tactics used by bad actors.
How do I integrate the WordPress plugin?
Sign up, get your API key from the developer dashboard, download the Guardian ChildSafe plugin ZIP, upload via Plugins → Add New → Upload Plugin, activate it, and paste your API key in Settings → Guardian ChildSafe. Your API key is also your plugin license key.
What is the typical API response time?
Median latency is 147ms for text and 200–300ms for image and video. We operate across 6 global regions to minimise round-trip time. Our SLA on Scale plans guarantees sub-300ms for 99th-percentile requests.
Can I try Guardian before committing to a plan?
Yes — all plans include a 14-day free trial with no credit card required. The free tier includes 1,000 API calls per month with access to all 35+ harm categories.

Start protecting children today

Join 500+ platforms that trust Guardian to keep their young users safe. Set up in 5 minutes. No credit card needed.

No credit card required Live in under 5 minutes COPPA & GDPR Ready Cancel anytime