Video intelligence agent for reasoning across entire video and image libraries.
Features
- Cross-library semantic search
- Multi-step video reasoning
- Person and scene tracking
- Automatic tagging and classification
- Highlight reel creation
- MCP, API, and SDK access
Use Cases
- Ad creative analysis
- Creator archive search
- Sports video research
- Video knowledge APIs
FAQ
Jockey uses TwelveLabs video-understanding models to reason across entire video and image knowledge stores. It searches people, scenes, quotes, and moments in natural language, tags and classifies media, discovers cross-library patterns, and assembles edit-ready highlights through Claude MCP, APIs, and SDKs. Core capabilities include: Cross-library semantic search, Multi-step video reasoning, Person and scene tracking.
Common scenarios include: Ad creative analysis, Creator archive search, Sports video research.
