I'm excited to have played a very small part in this release. Marengo 3.0 represents a step function change in multimodal understanding over extremely large corpus of data (petabyte++ scale). It unlocks enterprise & government use cases for semantic search and reasoning over videos.
Marengo 3.0 represents a fundamental shift from benchmark optimization to production-grade video understanding. By natively supporting multilingual content, composed queries, hour-long videos, and specialized domains like sports—while maintaining leadership on general benchmarks—we have built a foundation model that scales with real-world complexity.
reply