LynxVizion API

Process, analyze, and search video content with 15+ AI analysis modules.

Overview

LynxVizion is an AI-powered video analysis platform. Upload a video, select analysis modules, and receive structured, timestamped insights — visual elements, transcription, brand safety, scene detection, object tracking, and more.

Base URL:

https://api.lynxvizion.com/api

Key Capabilities

Video Processing

Upload videos and run 15+ AI analysis modules including visual analysis, transcription, brand safety, and scene detection.

Semantic Search

Search across all analyzed videos using natural language queries powered by vector embeddings (OpenAI text-embedding-3-small).

AI Chat Agent

Ask questions about any analyzed video. The AI agent uses full analysis context with BERT-compressed segment data.

HLS Streaming

Stream video playback via HLS with presigned B2 URLs. Includes thumbnail generation and direct MP4 fallback.

Analysis Modules

LynxVizion processes videos through these AI modules:

Module	Description
`visualelements`	Visual elements, objects, colors, composition
`deepvisualanalysis`	Deep frame-by-frame visual analysis (GPU)
`transcription`	Speech-to-text via Faster Whisper (GPU)
`transcription_diarization`	Speaker identification and diarization
`transcription_emotion`	Emotion detection in speech
`transcription_sentiment`	Sentiment analysis of transcribed text
`audio`	Audio event detection and classification
`scenes`	Scene and shot boundary detection
`locations`	Location and setting identification
`logos`	Logo and brand detection
`text`	On-screen text / OCR extraction
`brandsafety`	Brand safety assessment
`celebrities`	Celebrity recognition
`profiling`	Subject profiling and demographics
`body_language`	Body language and gesture analysis
`item_displacement`	Object movement and displacement tracking
`video_reasoning`	Temporal reasoning with YOLOv8 + CLIP

Authentication

All authenticated endpoints require an Authorization header with either:

API Key: Authorization: lvz_live_...
JWT (Clerk): Authorization: Bearer <token>

See the Authentication page for details.

Rate Limits

Usage is tracked per account with a default allocation of 3,000 processing minutes (50 hours). Additional minutes can be granted as bonus allocation. Check your current usage via the GET /usage/summary endpoint.