Accessibility in Live
- Verbit debuted Captivate Live at NAB, a product offering real-time captioning with low latency and high accuracy for broadcasts. - The demo positioned captions as suitable for live events, broadcasts, and streaming without introducing noticeable delay. - Presenters argued captions must be designed into live pipelines as a real-time data plane, not added after stream assembly. (sportsvideo.org)
Live captions are moving deeper into the broadcast chain, with Verbit using the 2026 NAB Show to debut Captivate Live for real-time production. (sportsvideo.org) At the Las Vegas show, Sports Video Group reported that Verbit pitched the product for broadcasts, live events and streaming, with an emphasis on low latency and high accuracy. NAB said its 2026 show ran April 18-22, with exhibits open April 19-22. (sportsvideo.org) (nab.org) Live captioning turns speech into on-screen text as people are still talking, so delays of even a few seconds can make a game, newscast or stream feel out of sync. Verbit’s Captivate Live listing on Amazon Web Services says the system is built for low-latency, high-accuracy captions and transcripts in fast-paced, multi-speaker environments. (aws.amazon.com) Verbit and NAB speakers framed captions as part of the live data flow, not a file added after the program is assembled. NAB’s 2026 conference agenda described live production as shifting from “video-only” workflows to systems that carry real-time metadata and accessibility layers alongside audio and video. (sportsvideo.org) (nab26.mapyourshow.com) That approach fits a wider push in media operations to treat captions as production infrastructure, especially for sports, news and streaming where viewers are watching live and distributors need text immediately. Verbit’s media product page says Captivate is designed for network news, live sports, streaming shows and independent content, with both live and post-production options. (verbit.ai) The company has been building toward this launch for more than a year. Verbit introduced Captivate in April 2024 as its in-house automatic speech recognition engine, and in 2025 it added speaker identification for live broadcast captions. (newscaststudio.com) (tvtechnology.com) Verbit says Captivate can be trained on customer terminology, formatting and industry-specific language, using term boosting and a dynamic dictionary to improve recognition. In sports, that matters for player names, team jargon and sponsor reads that generic speech engines often miss. (verbit.ai) (youtube.com) The company is also packaging captioning with controls and APIs aimed at production teams rather than only compliance staff. Verbit’s developer docs list live booking and caption control APIs, and the company’s support materials describe live workflows across Zoom, YouTube, conferences and streamed events. (verbit.readme.io) (verbit.my.site.com) At NAB this year, accessibility was being discussed alongside translation, audio description and live metadata, not as a separate afterthought. The result is a clearer pitch to broadcasters: if captions have to arrive in real time, they have to be built into the live stack from the start. (nab26.mapyourshow.com) (sportsvideo.org)
