Pegasus 1.5 by TwelveLabs
AI model for transforming video into Time-Based Metadata

Our Take
This is TwelveLabs doing what they do best — turning video into searchable, time-stamped metadata instead of just dumping a generic embedding. Pegasus 1.5 can apparently pinpoint specific moments, actions, and events in video with timestamps, which solves a real problem for anyone who’s ever had to scrub through hours of footage manually. I don’t have enough signal on adoption or accuracy to call it a winner yet, but the vertical they’re attacking — video intelligence at the moment level — is one of the few AI spaces that actually feels inevitable.
Transforms video into queryable, structured data assets using multimodal AI models to search, analyze, and generate text from video content at scale
Key Facts
The people behind Pegasus 1.5 by TwelveLabs
Links
Similar products worth knowing

IgnitionRAG
Multimodal RAG platform, from POC to production in minutes

Open Wearables
Open infrastructure for wearable-powered health products

Tabstack
Extract web data and automate browsers, no scraper required.

YourMemory
AI that remembers and forgets like humans.
Want products like this in your inbox every morning?
Five products. Every morning. Written by someone who actually cares whether they're good or not. Free forever, unsubscribe whenever.