The AI video discovery company Moments Lab has announced that it will unveil MXT-2, the latest version of its award-winning AI indexing technology, during the 2025 NAB Show.
The MXT-2 solution will be on display at the Las Vegas Convention center between April 6 to 9 at the company’s booth, SL 12213.
MXT is a multimodal and generative AI that understands what’s happening in every moment of a video and describes it like a human, saving creative teams hours of manual searching or scrubbing through videos, the company explained.
Through AI-generated, time-coded metadata, MXT not only enhances video search but can also generate customizable indexing that recognizes specific people, places, and logos and detects relevant moments depending on the content type.
MXT-2 uses a smarter, larger AI model than the previous version and is powered by more carefully curated data points for even more accurate, humanl-ike descriptions. MXT-2 is available on the Moments Lab platform or via its public API.
“We’ve designed MXT-2 to provide organizations with deeper editorial insights and video searchability at scale,” said Philippe Petitpont, co-founder and CEO at Moments Lab. “At NAB this year, we’re particularly excited to unveil the latest features added to our patented AI indexing technology, which is built using RAG (retrieval-augmented generation). RAG maximizes content creation, reuse, and repurposing, delivering tangible ROI by enabling entire media libraries to be described in natural language and therefore searched in this same language.”
More specifically, Moments Lab solutions available for demo at the 2025 NAB Show enable organizations to:
- Index and sort videos automatically with AI.
- Discover and repurpose content at scale.
- Share and distribute media files to key partners.
- Commercialize media libraries via a cloud-based storefront.
- Moments Lab to Present at 2025 NAB Show BEIT Conference
At this year’s NAB Show, Khalil Guetari, AI software engineer at Moments Lab, will give a presentation at the BEIT Conference (April 5-8) that will explore “Vision and Language Models for Enhanced Archive Video Management.” The session will take place on Monday, April 7, 10:40 AM - 11:00 AM PDT, in Room W222-W223. More information can be found here: nab25.mapyourshow.com/8_0/sessions/session-details.cfm?scheduleid=162.
