TwelveLabs Unleashes the Full Power of Video Intelligence With Jockey’s Public Debut
After completing its research preview with some of the world’s biggest content producers, new features bring video
Press Release Disclaimer: This is a press release distributed through the XPR Media network. It has not been independently verified by our newsroom.
SAN FRANCISCO, Oct. 15, 2026 (GLOBE NEWSWIRE) — TwelveLabs, the leading video intelligence company, today announced the official launch of Jockey by TwelveLabs. Jockey is TwelveLabs’ new video intelligence agent, serving as the orchestration layer that connects the company’s Marengo and Pegasus models; it brings video search and reasoning capabilities together in one unified workflow, creating an unparalleled experience for anyone working with high volumes of video data, from enterprise team to developers to creators.
Jockey lets users easily interact with their entire video libraries through the Claude or ChatGPT apps or build custom integrations via API. Using natural language, Jockey finds the moments that matter and explains what makes them relevant. It then acts on those results, depending on what a user needs. For example, Jockey can rank clips based on performance and assemble a rough cut video from source footage. Rather than requiring users to search, analyze, and organize footage manually, Jockey does it all and takes the next step forward.
With this release, Jockey also becomes the search layer for Marengo 3.5, the company’s latest embedding model, and draws on the recently released Pegasus 1.6. Together, the three form a single stack: Marengo finds it, Pegasus understands it, and Jockey acts on it.
“Jockey is the connective tissue between our models that will take video understanding to the next level. It has been rigorously tested by a number of leading companies as well as prolific creators to ensure that it delivers on its promise,” said Jae Lee, CEO and co-founder of TwelveLabs. “Now, with one incredibly powerful, cohesive video intelligence system, video shifts from a useful asset to something truly transformative for anyone who has large stores of content.”
From Research Preview to Production
Built with input from content producers who depend on massive video archives, TwelveLabs added a new set of capabilities to Jockey, including:
- A model that operates where people already work. Available in the ChatGPT and Claude apps, Jockey preserves search context across platforms. For example, a producer can start a search in the morning in one app and pick it up that afternoon in the other, without losing the thread.
- Social video and trend intelligence. Jockey can bring in videos directly from a URL in YouTube, TikTok, and Instagram Reels to surface trends relevant to a creative brief. This enables creators and marketing teams to see what’s working on the platforms that matter to them, without hours of manual scrolling.
- Performance insights alongside footage. Transcripts, technical metadata, creator details, and engagement metrics are captured with each video. This means results can be ranked by how a clip performed, not just by what it contains.
- Results that teams can keep and share. Whether it’s an mp4 assembled from raw footage, insights and analysis from social content, interactive video-driven websites, or other artifacts, Jockey’s output can be saved, shared, and revoked, all backed by playable, grounded video citations.
“Jockey changes what our creative and marketing teams can take on,” said Adam DeGregorio, VP, Creative Director, Hims & Hers. ”It helps us search and select footage and analyzes how our content performs across campaigns so we can see what resonates and build on it. That saves us time and helps us be more ambitious and intentional in the stories we tell.”
Jockey’s release comes on the heels of the Pegasus 1.6 release last week, which expands TwelveLabs’ reasoning capabilities for Physical AI. Pegasus adds the ability to understand egocentric video so that machines can train on human experiences and multimodal data to improve how they interact with their environments.
Information on how Pegasus 1.6 enhances Physical AI is here. To learn more about how Jockey works, please visit here.
About TwelveLabs
TwelveLabs makes video one of an organization’s most useful assets. The company has created a full-stack video intelligence platform, featuring its video AI agent Jockey alongside its industry leading video understanding models Marengo and Pegasus, designed to see and understand video the way humans do in a fraction of the time. Developers, enterprises, and creatives can now leverage this single system that gets smarter over time to access and act on all of their video content. TwelveLabs is backed by leading investors, including Amazon, Index Ventures, Naver, NEA, and Radical Ventures. For more information, visit www.twelvelabs.io.
Media Contact: Amber Moore amber@moorecom2.com
Media gallery
