Two endpoints report the state of each moderation type and flag. They need no API key, so you can call them from monitoring tools.
Current status#
curl https://api.detectivision.ai/api/v1/status{
"text_moderation": { "status": "operational", "flags": { "bad_habits": "available", "hate": "available", "insult": "available", "profanity": "available", "racism": "available", "self_harm": "available", "sexism": "available", "sexual": "available" } },
"image_moderation": { "status": "operational", "flags": { "hate": "available", "scam": "available", "sexual": "available", "violence": "available" } },
"video_moderation": { "status": "operational", "flags": { "alcohol": "available", "brand": "available", "explicit_nudity": "available", "nudity": "available", "sexual_objects": "available", "smoking": "available" } },
"audio_moderation": { "status": "operational", "flags": { "hate": "available", "insult": "available", "profanity": "available", "self_harm": "available", "sexual": "available" } },
"live_stream_moderation": { "status": "operational", "flags": { "explicit_nudity": "available", "nudity": "available", "sexual_objects": "available", "smoking": "available" } },
"updatedAt": "2026-10-16T02:56:26.869Z"
}status |
Meaning |
|---|---|
operational |
Every flag of the type can be checked. |
degraded |
Some flags cannot be checked right now (they are unavailable in flags). Checks run and return partial results, which are free. |
unavailable |
No check of this type can run right now; requests get 503 with Retry-After. |
coming_soon |
The type cannot be served right now; requests for it get 503 with not_available_yet. |
The answer is cached for 15 seconds.
History#
curl https://api.detectivision.ai/api/v1/status/historyThe last 24 hours, hour by hour, per type:
{
"hours": ["2026-09-26T11:00:00.000Z", "2026-09-26T12:00:00.000Z"],
"types": {
"image_moderation": { "uptime": 99.93, "hourly": [1, 0.983] }
},
"updatedAt": "2026-09-27T10:15:00.000Z"
}hourshas 24 entries, the start of each hour, oldest first (shortened above).hourlyhas one value per hour: the share, from 0 to 1, of health checks that passed;nullwhen nothing was checked.uptimeis the percentage (0 to 100) over the 24 hours.
Health checks run every minute. The answer is cached for 60 seconds.
Using status in your integration#
- You do not need to check status before each request: a
503or apartialresult already tells you. Status is for dashboards, alerts and deciding when to retry. - The dashboard shows the same information under Service status.