Get started

Compared to Video Understanding

TwelveLabs offers two products for working with video. They use the same ingestion flow but support different reasoning scopes.

  • Video Understanding: Search, analyze, and embed individual videos with dedicated APIs.
  • Video agents (Jockey): Organize, track, and extract insights across video collections through a single API.
  • Use both together: Combine Video Understanding for embeddings and per-video analysis with Video agents (Jockey) for reasoning across video collections.

Detailed comparison by task

TaskVideo UnderstandingVideo agents (Jockey)
Search for moments in videos✅ Search API✅ Responses API
Summarize a single video✅ Analyze API⚠️ Not recommended - use Analyze API
Organize videos by topic❌✅ Responses API with structured output
Track entities across videos❌✅ Responses API
Build agent workflows❌✅ Multi-turn sessions + structured output
Generate vector embeddings✅ Embed API❌

Video agents (Jockey) limitations

  • Video rendering and editing: The platform returns text, data references, and timestamps, not processed video. Use the references in your own rendering pipeline.
  • Single-video analysis: For per-video analysis, use the Analyze API. The Video agents (Jockey) works best when reasoning across multiple videos.
  • Vector embeddings: Use the Embed API to generate embeddings.

Next steps