One unmoderated post
can kill your startup

CSAM. Hate speech. Harassment.
Every platform with user-generated content faces these threats.
Get enterprise-grade protection before disaster strikes.

api.vettly.dev

POST /v1/check/fast

Analyzing...

"action": "allow"

→ Safe to publish

Get Free API Key See Pricing →

Production-ready

Fast response times

Simple API

Custom policies

See it in action

Test real content against our AI moderation. Click a sample or enter your own.

Sample Messages

Result

Click a sample or enter content

REAL-TIME STREAMING

Watch content flow through your safety system

Sub-500ms moderation for live chat, AI chatbots, and streaming platforms. WebSocket and SSE support included.

Live Moderation Stream

WebSocket Connected

streaming.ts

import { createStreamingClient } from '@vettly/sdk';

const streaming = createStreamingClient('your-api-key');

const ws = streaming.connectRealtime({
  policyId: 'chat-policy',
  onResult: (result) => {
    if (result.safe) showMessage(result);
    else logBlocked(result);
  }
});

await ws.connect();
await ws.moderate(message); // <500ms

Live Chat

Moderate messages before they appear in your chat feed

AI Chatbots

Screen LLM outputs in real-time before showing to users

Streaming

Flag uploaded content as frames are processed

Gaming

In-game chat moderation with zero perceptible latency

<500ms

Average latency

WebSocket

Persistent connection

SSE

Progress streaming

Get Free API Key

Learn from their mistakes

These platforms ignored moderation. Now they're case studies in what not to do.

Parler

Dead in 30 Days

What happened:

January 6th Capitol riot planned on platform

Moderation failure:

No proactive detection, manual review only

Result:

Deplatformed by AWS, Apple, Google in 48 hours

LESSON:

Reactive moderation = platform death

Voat

Toxic Brand

What happened:

Became haven for hate speech, no moderation

Moderation failure:

"Free speech absolutism" = no moderation

Result:

Shut down after 7 years, zero revenue

LESSON:

No users want unmoderated platforms

Jailbait Crisis

What happened:

Child exploitation subreddit went viral on CNN

Moderation failure:

Relied on community self-moderation

Result:

Nearly lost all advertisers, emergency overhaul

LESSON:

Even giants aren't immune

Don't become a cautionary tale.

Get protected before launch. Every platform that ignored moderation regretted it.

Get Your API Key

WHAT YOU GET

Everything you need

A complete moderation platform, not just an API.

Text, Images & Video

One unified API for all content types. Same request format, same response schema.

Custom Policies

Define rules in YAML. Set thresholds per category.

Real-Time Streaming

Sub-500ms WebSocket moderation for live chat and AI chatbots.

Webhooks

Get notified when content is flagged or blocked.

Dashboard

Monitor decisions, review trends, export logs.

Deterministic

Same input, same output. Every time. Predictable JSON with clear actions and confidence scores.

AI Agent Ready

Works with Claude, Cursor, and any MCP-compatible AI tool. One config line.

Get Free API Key

POLICIES IN PLAIN ENGLISH

Rules you can actually read

Define moderation rules in simple YAML. No complex regex, no black-box algorithms. Just human-readable policies that do exactly what they say.

marketplace-safe.yml

name: marketplace-safe
description: Protect buyers and sellers

rules:
  - category: violence
    threshold: 0.3
    action: reject

  - category: hate
    threshold: 0.3
    action: reject

  - category: scam
    threshold: 0.5
    action: flag
    customPrompt: "Does this show counterfeit luxury goods?"
    customCategory: counterfeit_detection

Block violent content

Reject images or text with violence scores above 30%

30%

No hate speech

Reject content with hate speech above 30% confidence

30%

Detect counterfeits

Pro+

AI answers: "Does this show counterfeit luxury goods?"

50%

Custom Prompts

Pro+

Ask AI anything about your images. Write rules in plain English like "Is this counterfeit?" or "Does this food look undercooked?"

customPrompt: "Does this show counterfeit luxury goods?"

Start with a template

Marketplace

Scams & counterfeits

Gaming

Toxicity & threats

Education

Strict for minors

Social

Balanced community

Custom

Build your own

ENTERPRISE-GRADE

Advanced features included

These normally cost $50K+ to build. You get them out of the box.

Appeals Workflow

Built-in user appeals with admin review system. Let users contest decisions with full audit trails.

Evidence Storage

GDPR-compliant S3/R2 storage for flagged content. Automatic retention policies and secure access.

Bot Detection

Behavioral signals, account age analysis, and coordinated campaign detection built-in.

Audit Trails

Export compliance reports with complete decision history. Perfect for legal and regulatory needs.

CSAM Reporting

Automatic reporting to NCMEC for child safety violations. Enterprise-grade compliance.

Circuit Breakers

Automatic failover when providers are down. Rate limit handling and retry logic included.

Multi-Provider Fallbacks

OpenAI, Hive, Azure, Perspective - all integrated with automatic provider switching.

Video Frame Extraction

FFmpeg-powered frame analysis with BullMQ job queues. Async processing for scale.

These features are typically custom-built for enterprise contracts. With Vettly, they're included in every paid tier—even Starter at $29/mo.

BUILT WITH VETTLY

From zero to production in minutes

We built this Discord bot with Vettly. Add it to your server or build your own.

Vettly Bot

Online

BadActorToday at 2:34 PM

I hate everyone in this server you all suck

Vettly BotBOTToday at 2:34 PM

Message blocked

Detected: harassment, toxic content

GoodUserToday at 2:35 PM

Hey everyone! Just finished the tutorial, this is awesome 🎉

Vettly BotBOTToday at 2:35 PM

Message approved

Protected

Real-time

One API call. That's it.

This bot took an afternoon to build. The moderation logic? Just one line of code. Try it on your server or check out the source to build your own.

// The entire moderation logic

const result = await vettly.check({ content, contentType: "text" })

Full Discord bot in ~200 lines
Text and image moderation
Configurable policies per server
Open source on GitHub

Add to Discord View Source

DEVELOPER EXPERIENCE

Ship in minutes

One API call. Consistent response format. That's it.

Request

curl -X POST https://api.vettly.dev/v1/check \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{"content": "Text to moderate", "policy": "default"}'

Response

{
  "action": "block",
  "categories": {
    "hate": 0.91,
    "harassment": 0.08,
    "violence": 0.12
  },
  "flags": ["hate"],
  "latency_ms": 147
}

DASHBOARD

Everything at a glance

Real-time analytics, decision logs, and policy performance. All in one place.

vettly.dev/dashboard

Today's Checks

+23% from yesterday

Blocked

2.7% block rate

Avg Latency

0ms

Under SLA

Accuracy

Based on appeals

Request Volume

Allowed

Blocked

00:0006:0012:0018:00Now

Recent Decisions

Text

Great product, highly recommend!

allow

Image

profile_photo_2847.jpg

allow

Text

You're worthless and should...

block

Text

This might be spam...

flag

Simple, honest pricing

Start with a limited Developer plan. Paid tiers fund infra and higher limits.

Moderation costs money to run - especially at scale and across video.
We keep a limited Developer plan and price the rest to be sustainable.

Developer

Limited usage for evaluation

$0forever

2,000 text checks / month
100 image checks / month
25 video checks / month
Hate, violence, sexual, self-harm detection
Dashboard & 24-hour logs
Pre-built policies

Get Free API Key

Growth

Ship publicly or submit to App Store

$49/month

50,000 text checks / month
5,000 image checks / month
1,000 video checks / month
Detects spam, scams, and abusive content
Blocks App Store rejection content
Webhooks to flag, block, or review
7-day moderation logs

Pro

Users, payments, or legal exposure

$149/month

250,000 text checks / month
25,000 image checks / month
5,000 video checks / month
Everything in Growth
Custom moderation rules
Priority processing
30-day audit logs for compliance
Protects Stripe, Apple, Google Play

Enterprise

Moderation failure would harm your business

Custom

Everything in Pro
Batch moderation API
Analytics, exports, and reporting
Dedicated support
SLA & uptime guarantee

Contact Sales

All plans include: API access, dashboard, real-time moderation, and our unified text/image/video endpoint.

Consistent API

Same JSON structure for every request. Text, images, video — one predictable format.

Built for Speed

Optimized for low-latency responses. See real performance in the playground above.

Your Data, Your Control

We don't train on your content. Configure retention policies to match your requirements.

Production Infrastructure

Deployed on enterprise-grade cloud infrastructure with monitoring and alerting.

One unmoderated postcan kill your startup

Recent Content Moderation Failures

X (Twitter)

Meta

X (Twitter)

Meta

TikTok

Apple

Snapchat

Discord

Meta

TikTok

X (Twitter)

Meta

X (Twitter)

Meta

TikTok

Apple

Snapchat

Discord

Meta

TikTok

See it in action

Sample Messages

Watch content flow through your safety system

Live Chat

AI Chatbots

Streaming

Gaming

Learn from their mistakes

Parler

Voat

Reddit

Everything you need

Text, Images & Video

Custom Policies

Real-Time Streaming

Webhooks

Dashboard

Deterministic

AI Agent Ready

Rules you can actually read

Block violent content

No hate speech

Detect counterfeits

Custom Prompts

Start with a template

Advanced features included

Appeals Workflow

Evidence Storage

Bot Detection

Audit Trails

CSAM Reporting

Circuit Breakers

Multi-Provider Fallbacks

Video Frame Extraction

From zero to production in minutes

One API call. That's it.

Ship in minutes

Everything at a glance

Simple, honest pricing

Developer

Growth

Pro

Enterprise

Consistent API

Built for Speed

Your Data, Your Control

Production Infrastructure

One unmoderated post
can kill your startup