Content moderation API

We return numbers.
You decide the policy.

Submit text, an image, a GIF or a video and get a score per category — toxicity, harassment, hate, violence, sexual content, gore. No thresholds, no verdicts, no opinion about what is acceptable on your platform.

POST /moderations

{
  "status": "completed",
  "scores": {
    "toxicity":   0.9889,
    "harassment": 0.7589,
    "violence":   0.7812,
    "profanity":  0.6626,
    "hate":       0.0432,
    "sexual":     null,
    "gore":       null
  }
}

0.0432 means a model looked and found almost nothing. null means nothing looked — that category was not assessed for this content type. Treating null as safe is the one mistake that will bite you.

Plans

Free

$0/ month

Enough to wire up an integration and calibrate your thresholds.

50 production requests / day

Start on Free

Growth

$19/ month

For a live product with real moderation volume.

50,000 production requests / day

Start on Growth

Scale

$49/ month

High-volume moderation with headroom for spikes.

500,000 production requests / day

Start on Scale

Every plan includes a sandbox and a production environment, with separate keys and separate quotas: sandbox is metered as the free tier — 50 requests a day, whatever you pay — so testing never spends what live traffic is paying for. Signed webhooks on every plan.