Generative AI has made it incredibly easy to churn out new videos, but finding a specific moment buried inside years of old footage remains a frustrating, manual chore.
Clipto, a startup building systems to solve this unstructured data problem, announced today it has raised $15 million in funding at a $250 million post-money valuation. Backed by a roster of global investors including HSG (formerly Sequoia China), GL Ventures, EnvisionX Capital, and Palm Drive Capital, the company is building what it calls an AI memory layer for video, allowing both human users and AI agents to instantly search private media libraries.
The current standard for video search is largely archaic. Creators, enterprise teams, and casual users are capturing terabytes of footage, only for it to sit stagnant on hard drives. Finding a specific clip usually involves manually scrubbing through timelines, decoding cryptic file names, or spending hours tagging files. While cloud-based AI tools exist to help, they require uploading massive video files, creating bandwidth bottlenecks, driving up costs, and introducing significant privacy concerns.
Clipto bypasses the cloud entirely. By utilizing proprietary on-device AI, the platform indexes media locally. It combines semantic search, automatic transcription, and structured notes to turn disconnected footage into a highly searchable database without compromising privacy. A user can simply type a natural language prompt and jump straight to the exact timecode of the shot they need.
The approach is already gaining massive traction. Founded just last year in 2023, Clipto claims its platform is used by more than 30 million people worldwide, including professionals at tech giants like Apple, Meta, Google, and Microsoft, as well as enterprise brands like McKinsey and L’Oréal. Alongside the funding announcement, Clipto is expanding its footprint by launching native Windows and Android applications to complement its existing Web, Mac, and iOS clients. Deep integrations with industry-standard tools like Adobe Premiere also allow editors to cut hours out of their post-production workflow.
Perhaps the most forward-looking update is Clipto’s new support for the Model Context Protocol (MCP). This integration allows external AI agents to securely query a user’s private, local video library. Instead of just relying on web data or text documents, AI systems can now scrub through personal media archives to find and analyze relevant footage, giving AI agents a secure window into user-generated video content.
“Generative AI has transformed how content is created, but it still struggles to understand the enormous amount of personal media people already have,” said Henry Kang, Founder and CEO of Clipto. Kang argues that the next frontier of AI requires a dedicated memory layer to complement its reasoning capabilities. By keeping data processing strictly on-device, Clipto is positioning itself to unlock years of previously inaccessible footage for the modern AI ecosystem.
Try Clipto today at https://www.clipto.com/.
