Starter
$99$199 per month, billed yearly $1,188per month, monthly
3 reviewer seats, 100,000 items a month through AI scoring.
Choose StarterAI content moderation that knows when to ask a person
Models are fast and cheap on clear cases and unreliable on context. We use AI for what it does well and put trained people on everything else, with thresholds you approve.
Queue 4
Profiles, Spanish
Reports, English
Listings, Portuguese
Lucía, 31
Madrid, joined today
Nurse, love the sea and salsa nights.
Not here much, write me on Telegram @lu_31
Policy
Romance scam pattern
Section 4.2 Moving users off the platform to ask for money
AI score
0.71
Audit log
09:40MD-48210 AI 0.55 → human review → Approved, reviewer 022
09:39MD-48209 AI 0.97 → removed under your threshold
09:38MD-48207 AI 0.71 → human review → Removed, reviewer 014
Where automated content moderation works
- Spam, duplicates and copy paste abuse at high volume
- Explicit nudity and clear graphic violence in images
- Known prohibited terms, links and contact details in listings
- Language detection and routing to the right reviewer
- Ranking a queue by risk so the worst items are seen first
Where models still need people
- Context and intent: a slur quoted in a report, sarcasm, reclaimed words
- Romance scams and fraud that look polite message by message
- New slang, coded language and emoji used to evade filters
- Local law and cultural norms that differ by market
- Appeals, where the user adds context the model never saw
- Generated content that imitates real people or brands
Two thresholds and three outcomes
Every item gets a score per policy category. Above the upper threshold the item closes automatically. Below the lower threshold it passes. Everything between the two goes to a moderator.
You approve both thresholds per content type. Tighter thresholds mean more human review and fewer automatic mistakes. Looser thresholds mean less cost and more automatic decisions. The weekly report shows the split and the QA results, so the trade off is a decision you make with data.
The model is checked too
QA samples automatic decisions every week, not only human ones. When a model drifts after an update or a new abuse pattern appears, people see it before your users do.
AI scoring by content type
See also image moderation, video moderation and NSFW detection.
| Content type | What AI scores | What people decide |
|---|---|---|
| Text and comments | spam, toxicity, threats, personal data | context, intent, quoted speech |
| Chat and DMs | scam patterns, harassment, links | conversation history, grooming signals |
| Images | nudity, violence, symbols, text in image | borderline nudity, art, news value |
| Video | frames and audio flagged for review | context across the whole clip |
| Profiles | stolen photos, contact details, impersonation | real identity, verification disputes |
| Listings | prohibited items, counterfeit terms | price anomalies, fraud patterns |
See your AI and human split
The plan builder estimates the share AI resolves for your content mix and the human team you need for the rest.
Get your moderation plan in two minutes
Tell us your platform type, content, volume, languages and SLA. The plan shows your team per shift, the AI and human split and a monthly estimate.
Get my moderation plan