Video-native AI for understanding, searching, and generating text from video content.
Free plan available, paid tiers aboveVisit WebsiteFreemiumVisit Website
Tracked since2026
The Bottom Line
- Price
- Free plan available, paid tiers above. Compare plans
Key facts
- Video-native AI models that understand video holistically, integrating visual, audio, and speech information for superior search and analysis.
- Enables semantic search, text generation (summaries, chapters, highlights), and custom classifier building using natural language.
- Specialized solution for Physical AI/robotics to turn first-person video into structured training data with action segmentation, quality scoring, and compliance flagging.
Pros
- Video-native AI provides deeper understanding than frame-by-frame or transcription-only approaches.
- Supports multiple use cases from media search to robotics training data preparation.
- Offers flexible deployment and integration via API and SDK.
Cons
- Pricing details are not transparent on the website, requiring a sales conversation.
- May require technical expertise to fully leverage the API and SDK for custom pipelines.
What is Pegasus 1.6 by TwelveLabs?
TwelveLabs provides a video AI platform that uses video-native foundation models to understand, search, analyze, and generate text from video content. The platform offers two main models: Marengo, a multimodal embedding model for any-to-any retrieval across video, audio, images, text, and documents; and Pegasus, a video-first language model for generating descriptions, summaries, and analysis. It enables semantic search, video-to-text generation, and custom classifier building through embeddings. The platform is designed for industries including media, advertising, government, security, automotive, and physical AI/robotics, where it helps convert raw video into structured, training-ready data.
Key Features
- Semantic search across video using natural language queries
- Video-to-text generation including summaries, chapters, highlights, Q&As, and hashtags
- Video embeddings for building custom classifiers and recommendation systems
- Action segmentation and labeling for robotics training data
- Quality scoring for video clips based on stability, framing, occlusion, and action clarity
- Consent and compliance flagging to detect PII and minors in footage
- Dense captioning of hand-object interactions and spatial relationships
- Cross-modal search across text, audio, image, video, and documents
Pricing Plans
Pricing checked Oct 8, 2026
| Plan | Price | Details |
|---|---|---|
| Free | Free |
+14 moreShow less
|
| Developer | Pay as you go |
+13 moreShow less
|
| Enterprise | Custom |
+7 moreShow less
|
Reviews
$99Free with your review
Write a reviewReview Pegasus 1.6 by TwelveLabs, get a free AI guide
Share your experience and we will send you Improve Your Thinking Patterns Using ChatGPT, free.
Best Pegasus 1.6 by TwelveLabs Alternatives
Top alternatives based on features, pricing, and user needs.
Still deciding?
Most buyers shortlist 2 or 3 tools before committing. Pull a side-by-side comparison or browse the full alternatives shortlist below.
Explore More
Pegasus 1.6 by TwelveLabs FAQ
How does Pegasus differ from Marengo?
Pegasus is a video-first language model that generates text descriptions, summaries, and analysis from video, while Marengo is a multimodal embedding model for any-to-any retrieval across video, audio, images, text, and documents.
Can I use the platform to build a custom video classifier?
Yes, the Embed capability allows you to build your own classifier using natural language and run it instantly, leveraging video embeddings.
Does the platform support searching across multiple video dimensions?
Yes, semantic search works across speech, text, audio, and visuals, providing context-aware results that understand all dimensions of the video.
What is the Physical AI solution for robotics teams?
It converts first-person video into structured training data by offering action segmentation, dense captioning, quality scoring, search and curation, and consent/compliance flagging, all customizable with your own taxonomies.
Can I integrate the platform with my existing data pipeline?
Yes, the platform provides an API and SDK with flexible deployment options, and outputs structured JSON with timestamped segments that fit into downstream workflows.
What types of text outputs can Pegasus generate?
Pegasus can generate summaries, chapters, highlights, Q&As, hashtags, marketing suggestions, captions, and domain-specific analysis from video content.
Is there a way to try the platform without commitment?
Yes, you can try the platform for free via the Playground, and there is a 'Try Free' option available on the website.
What industries does TwelveLabs cater to?
The platform serves media and entertainment, advertising, government and security, automotive, sports and broadcasting, and physical AI/robotics industries.
Source: twelvelabs.io