Introduction
TwelveLabs is a video intelligence platform. Upload your videos and use the API to search, analyze, generate embeddings, or reason across entire collections. You can access all capabilities through REST APIs or the Python and Node.js SDKs.
Single videos or collections of videos?
TwelveLabs provides two products: one for analyzing single videos and another for reasoning across entire collections.
Search for specific moments, analyze content, and generate embeddings, one video at a time, with dedicated APIs for each task.
Organize, search, track entities, and extract insights across your entire video collection through a single API. This feature is in private beta.
Not sure which option to use?
Single videos: Use video understanding
- Search for specific moments in a video using text or image queries
- Analyze an individual video
- Generate vector embeddings for machine learning pipelines
Collections of videos: Use video agents (Jockey)
- Search, analyze, and extract insights across an entire video collection
- Organize and classify a video library by topic
- Track entities across multiple videos
- Build multi-step agent workflows with structured output
- Refine results through follow-up requests in multi-turn sessions
Need both? Use Video Understanding for embeddings and analysis of individual videos, while utilizing the Video agents (Jockey) for corpus-level reasoning. Both products share the same method of ingesting content.
Already using Video Understanding? See how the two products compare and the migration guide.
