Video Intelligence MCP
Comprehensive video analysis: object detection, scene classification, activity recognition, and narrative summarization. Returns timestamped object tracks, scene labels, detected activities, and an executive summary of video content.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
video_url |
string | ✅ | Video file URL |
tasks |
array | ❌ | objects, scenes, activities, summary (default: all) |
confidence_threshold |
float | ❌ | Min detection confidence 0.0–1.0 (default: 0.7) |
frame_sample_rate |
integer | ❌ | Frames per second to sample (default: 1) |
Example Response
{
"duration_seconds": 98.3,
"objects_detected": [
{"label": "person", "count": 3, "first_seen": "0:02", "confidence": 0.97},
{"label": "laptop", "count": 2, "first_seen": "0:08", "confidence": 0.89}
],
"scenes": [
{"timestamp": "0:00–0:35", "label": "office_meeting", "confidence": 0.93},
{"timestamp": "0:35–1:38", "label": "product_demo", "confidence": 0.88}
],
"activities": ["discussion", "screen_sharing", "whiteboard_drawing"],
"summary": "Three-person product demo meeting. Presenter shows software on laptop. Whiteboard session at 0:52."
}
Use Cases
- Security camera automated incident detection
- Video content moderation at scale
- Sports video highlight clip identification
Access
# x402 endpoint — pay $0.05 USDC per call (Base mainnet)
POST https://x402.ntriq.co.kr/video-intel-mcp
# Service catalog
curl https://x402.ntriq.co.kr/services
x402 micropayments — USDC on Base, gasless EIP-3009