TwelveLabs

What if your AI could actually understand what’s happening in a video? Not just the objects on screen, but the context, tone, and narrative across time. This is where TwelveLabs comes in. In this short demo, we show how you can: 🔍 Search video using natural language ✨ Generate summaries, highlights, and insights 🧠 Go beyond keywords, toward real understanding From sports to security, media to enterprise ops, this is what’s possible when you unlock the full potential of video. #VideoAI #TwelveLabs

Thrilled to see TwelveLabs highlighted in Amazon’s Q2 earnings release! 🎉 And after just two weeks live on Amazon Bedrock, the inbound interest from builders and enterprises has been nothing short of extraordinary -- and we’re already well into shipping what’s coming next! By bringing our Marengo 2.7 and Pegasus 1.2 video-understanding models to Bedrock, we’re helping customers transform every second of their precious video assets into instantly searchable, actionable insight at global scale. A heartfelt thank-you to the visionary teams at Amazon and Amazon Web Services (AWS) who share our mission, and to every member of the TwelveLabs family whose passion and grit made this milestone possible. This partnership is only just beginning, and the road ahead is wide open. 🚀

In the high-stakes world of sports, delivering highlight-worthy content fast is everything. That’s why we’re teaming up with Snowflake to showcase how video AI + data cloud infrastructure is transforming the game. Join Travis Couture, Director of Solutions Engineering at TwelveLabs, and Michelene (Loomer) Rabbitt, Industry Principal at Snowflake, as they dive into how sports organizations can: 📊 Use Snowflake’s AI Data Cloud to scale and prep massive volumes of sports data 🎥 Leverage TwelveLabs models to understand game footage like a human, identifying plays, generating highlights, and more ⚡ Streamline workflows, cut manual effort, and accelerate time to story Whether you're building for fans, broadcasters, or internal ops, this is where the future of sports content begins. 🗓 August 7, 2025 ⏰ 10:00 AM PT Save your spot now: https://shorturl.at/HEpcE #TwelveLabs #VideoAI

This past weekend, TwelveLabs had the incredible opportunity to sponsor Hack the 6ix, Toronto’s largest summer hackathon. 🇨🇦 Over 400 talented hackers came together from across Canada, many from top-tier schools like University of Toronto and University of Waterloo, to build, collaborate, and push the boundaries of what's possible with AI. Out of ~130 projects, nearly 30 teams built with the TwelveLabs API, applying video understanding to everything from: 📚 Education & training 🎮 Gaming & entertainment 🔍 Discovery & search 🧠 Content creation 💬 Communication & analysis We’re blown away by the creativity and real-world potential we saw. Thank you to the amazing team at Hack The 6ix for putting together such an event, TwelveLabs’ James Le and Evan Toyne for representing, and to fellow sponsors Vellum, Linear, Warp, and Major League Hacking for supporting the next generation of builders. #TwelveLabs #VideoAI #Hackathon

🚨 Big news from the Amazon Web Services (AWS) Summit NYC stage! Our Co-Founders, Jae Lee and Soyoung Lee, went live with the AWS team to announce that TwelveLabs' video foundation models are now available in Amazon Bedrock. What’s possible now? 🔍 Retrieve the exact scene you’re looking for - fast! 🎭 Surface emotional moments, reactions, or gestures 🧠 Organize petabytes of footage into searchable knowledge This launch marks a major milestone in our mission to help organizations unlock the 90% of data that lives in video. And with TwelveLabs models now in Bedrock, you can get started in just a few clicks, no data movement required. Watch the full demo and walkthrough (starting at 45:00): https://lnkd.in/garUenmy #TwelveLabs #VideoAI

In the 89th session of #MultimodalWeekly, we have two exciting presentations from researchers working on reasoning models: ✅ Rulin Shao and Rui Qiao will present their work called ReasonIR - the first retriever specifically trained for general reasoning tasks, achieving the state-of-the-art retrieval performance on BRIGHT (reasoning-intensive retrieval). When employed for retrieval-augmented generation (RAG), ReasonIR-8B also brings substantial gains on MMLU and GPQA. - ReasonIR Paper: https://lnkd.in/gPbpq9Jw - ReasonIR Codebase: https://lnkd.in/ev2eKkZB ✅ Zhen Zhang (with Xin Eric Wang) will present his work called "Soft Thinking" - a training-free method that emulates human-like "soft" reasoning by generating soft, abstract concept tokens in a continuous concept space. In essence, each generated concept token encapsulates multiple meanings from related discrete tokens, implicitly exploring various reasoning paths to converge effectively toward the correct answer. - Soft Thinking Paper: https://lnkd.in/gVQxNDaQ - Soft Thinking Codebase: https://lnkd.in/geGspS5q Register for the webinar here: https://lnkd.in/gJGtscSH ⬅️

🔍 Searching through vast media libraries just got smarter. See how Amazon S3 Vectors (preview) & TwelveLabs on Amazon Bedrock are enabling natural language search across media assets - turning hours of searching into seconds of discovery. 📍 Learn more: https://go.aws/4nUmdi8 #AWSforME #AmazonBedrock #AmazonS3

What does it really take to build AI that sees, hears, and understands video like a human? At #AWSSummitNYC, our co-founder, Soyoung Lee, joined SiliconANGLE & theCUBE to break it all down. From the complexity of multimodal AI to real-world use cases like: 🔍 Searching decades of video footage in seconds ✂️ Instantly generating highlight reels 🛡️ Making sense of safety and surveillance footage 🧠 Powering personalized content and recommendations She also shared how developers and enterprises are using our API-first video foundation models, now available in Amazon Web Services (AWS) Bedrock, to turn massive video libraries into intelligent, searchable assets. Check out the whole conversation to hear why we think the future of video is multimodal: https://lnkd.in/grr3WMAN #TwelveLabs #VideoAI

Big news: TwelveLabs models are now in Amazon Web Services (AWS) Bedrock! 🎉 This marks a huge leap forward in how organizations can search, analyze, and extract insights across massive video libraries—securely and at enterprise scale. Think: 🎙️ Finding a specific moment across seasons of content ⚽ Locating every moment a player celebrates a goal 🎞️ Generating highlights and trailers in minutes, not weeks This isn't just faster search, it's deep video understanding with enterprise-grade infrastructure. We're proud to team up with AWS to help organizations tap into the 80% of data that's been largely invisible until now. Learn more here: https://lnkd.in/g_xcDRaH #VideoAI #TwelveLabs

Video search gets even smarter. 🎬🧠⚡️ https://go.aws/4lz4ueC Now available in Amazon Bedrock, TwelveLabs' fully managed, serverless models brings human-level video understanding to your applications. Search & analyze content through actions, objects & sounds, transforming video into precise insights. #AWS #AWSSummit