> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tuteliq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction

> Detect grooming, bullying, self-harm, fraud, and more with a single API call

## Ship child-safe apps without building content moderation from scratch

Tuteliq is a content moderation API aligned with KOSA. Detect grooming, bullying, self-harm, fraud, social engineering, and more — with age-calibrated risk scoring designed for near-real-time moderation. Verify user age and identity to prevent impersonation and ensure compliance. One API call replaces months of in-house ML work.

<CardGroup cols={2}>
  <Card title="Get API Key — Free" icon="key" href="https://tuteliq.ai/dashboard">
    Create your account and start analyzing content immediately.
  </Card>

  <Card title="Read the Docs" icon="book" href="/quickstart">
    Follow the quickstart guide to make your first API call.
  </Card>
</CardGroup>

<Info>
  Built for platforms where minors interact online — from gaming chat to classroom apps.
</Info>

## The problem

Child safety compliance shouldn't require a machine learning team. KOSA requires platforms to protect minors from nine categories of harm — bullying, grooming, eating disorders, substance use, self-harm, depression, compulsive usage, sexual exploitation, and unsafe visual content. Building detection for even one of these categories takes months of ML engineering, training data, and ongoing maintenance. Building for all nine, across text, voice, and images, with age-appropriate calibration? That's a team-year of work.

Tuteliq does it in a single API call.

<CardGroup cols={2}>
  <Card title="Safety Detection" icon="shield-halved" href="/quickstart">
    Per-message risk scoring for bullying, grooming, self-harm, substance use, and more — calibrated by age group and aligned to all nine KOSA harm categories.
  </Card>

  <Card title="Fraud & Scam Detection" icon="user-secret" href="/api-reference/introduction">
    Detect social engineering, app fraud, romance scams, and money mule recruitment targeting minors and vulnerable users.
  </Card>

  <Card title="Extended Safety" icon="shield-check" href="/api-reference/introduction">
    Gambling harm, coercive control, vulnerability exploitation, and radicalisation detection — with cross-endpoint vulnerability profiling.
  </Card>

  <Card title="Emotional Analysis" icon="brain" href="/api-reference/introduction">
    Track emotional trends over time to catch depression, anxiety, and declining mental health before they escalate — not just per-message, but across entire conversation histories.
  </Card>

  <Card title="Voice, Image & Video" icon="microphone" href="/voice-streaming">
    Process audio, images, and video for safety analysis — transcription, OCR, frame extraction, and visual content classification. Real-time streaming for live monitoring.
  </Card>

  <Card title="Batch & Webhooks" icon="arrows-spin" href="/webhooks">
    Submit bulk analysis jobs and receive results asynchronously via webhook callbacks.
  </Card>

  <Card title="KOSA Compliance" icon="gavel" href="/kosa-compliance">
    Built-in coverage for all nine Kids Online Safety Act harm categories with audit-ready reporting.
  </Card>

  <Card title="GDPR Ready" icon="lock" href="/gdpr">
    Data minimization, retention controls, and consent management designed for processing children's data.
  </Card>

  <Card title="Age Verification" icon="id-card" href="/verification">
    Verify user age through document analysis, MRZ parsing, barcode decoding, and biometric estimation — with 45-country document validation and multi-source cross-referencing.
  </Card>

  <Card title="Identity Verification" icon="fingerprint" href="/verification#identity-verification">
    Full identity verification with document authenticity analysis, face matching, visual liveness detection, 7-layer fraud cross-referencing, and recapture detection.
  </Card>
</CardGroup>

## Why Tuteliq

<CardGroup cols={2}>
  <Card title="Age-Calibrated Severity" icon="child">
    A joke between 16-year-olds isn't the same as a message to an 8-year-old. Risk scores automatically adjust across four age brackets so your moderation matches developmental context.
  </Card>

  <Card title="Context, Not Keywords" icon="brain-circuit">
    A teen texting "I'm literally going to die if I don't get those shoes" shouldn't trigger the same response as genuine crisis language. Tuteliq's context engine understands the difference — dramatically reducing false positives.
  </Card>

  <Card title="Beyond Detection" icon="clipboard-list">
    Most safety APIs stop at a risk score. Tuteliq generates age-appropriate action plans for children, parents, and moderators, plus professional incident reports ready for school counselors and compliance audits.
  </Card>

  <Card title="Full KOSA Coverage" icon="shield-check">
    Nine out of nine harm categories covered out of the box. No mix-and-match from multiple vendors. One integration, full compliance.
  </Card>

  <Card title="Multimodal & Multilingual" icon="layer-group">
    Text, voice, and images analyzed through a single API in 27 languages. Language is auto-detected — no configuration needed. Audio is transcribed and safety-analyzed with timestamped segments. Images are classified visually and OCR-scanned for embedded text.
  </Card>

  <Card title="Built for Production" icon="server">
    Predictable latency for near-real-time moderation (typical p95 under 1.5s for LLM-backed detection). 99.9% uptime SLA. Batch processing for up to 50 items per request. HMAC-signed webhooks with automatic retry. GDPR-compliant data management on every tier.
  </Card>
</CardGroup>

## Who uses Tuteliq

Built for platforms where minors interact:

* **Gaming platforms** — Moderate in-game chat, voice comms, and user-generated content in real time.
* **Social apps** — Detect grooming, bullying, self-harm, and romance scam signals in DMs and feeds before they escalate.
* **Financial platforms** — Flag social engineering, money mule recruitment, and app fraud targeting younger users.
* **Ed-tech** — Ensure safe learning environments with content filtering that understands classroom context.
* **Messaging apps** — Analyze conversations for emotional distress and predatory behavior across text and voice.

## Performance

| Metric                             | Value                                                                                      |
| ---------------------------------- | ------------------------------------------------------------------------------------------ |
| Typical p95 (LLM-backed endpoints) | \~1.4s                                                                                     |
| Uptime SLA                         | 99.9%                                                                                      |
| KOSA categories covered            | 9/9                                                                                        |
| Supported input types              | Text, voice, image, video                                                                  |
| Languages supported                | 27 — English (stable), all 24 EU official languages + Ukrainian, Norwegian, Turkish (beta) |
| Rate limits                        | 60–5,000 req/min depending on [tier](/authentication)                                      |
| Credit costs                       | 2–95 credits per call — see [Pricing & Credits](/credits)                                  |

## Get started

<Card title="Quickstart" icon="rocket" href="/quickstart" horizontal>
  Get up and running with Tuteliq in under 5 minutes.
</Card>