Cosmos raises $15M Series A to build a creative-first visual search engine and inspiration-driven commerce platform
Jan 20, 2026 with Andy McCune
Key Points
- Cosmos closes $15 million Series A for its visual search platform, which hosts tens of millions of user-uploaded images and runs hundreds of thousands of daily searches across creative teams at Nike, Chanel, and Apple.
- The company built proprietary aesthetic prediction models trained on user data to filter low-quality content and flags AI-generated images through detection algorithms, positioning itself against content quality deterioration.
- Cosmos is testing two monetization paths: an enterprise product for creative teams already at scale, and an inspiration-driven commerce pilot launching in 2025 that targets high-intent searches like 'brown leather couch' as natural purchase moments.
Summary
Cosmos, a New York-based visual search and inspiration platform, closed a $15 million Series A and launched a significant product update on the same day. The company, led by founder and CEO Andy McHughan, has been building for roughly four years following his exit from Unfold, a mobile content creation tool he founded in 2017 and sold to Squarespace.
The platform currently hosts tens of millions of images, runs hundreds of thousands of searches daily, and counts teams at Nike, Chanel, and Apple among its users. Growth has been entirely organic, with early traction built around a pre-launch waitlist that reached several hundred thousand users, driven primarily by designers across interior, graphic, brand, and architectural disciplines. The current growth wave is coming from Gen Z users who use the platform for mood boarding and aesthetic curation rather than professional reference work.
AI detection and content quality are central to Cosmos's positioning. Every uploaded image is run through multiple detection algorithms to flag AI-generated content, giving users the ability to show or hide it via profile settings. McHughan acknowledges the detection problem is an arms race that will only get harder. Beyond detection, the company has built proprietary in-house aesthetic prediction models trained on its own user data to re-rank search results and feed recommendations, designed to filter low-quality content rather than impose a house aesthetic.
Monetization is nascent. A premium subscription exists but is deliberately demoted within the product. The two strategic directions signaled are an enterprise or team-tier product targeting the creative teams already using the platform at scale, and an inspiration-driven commerce layer. A commerce pilot launched in 2025, with the thesis that high-intent searches like "brown leather couch" represent a natural purchase moment that can be served without a hard-sell product experience.
The product update released on the day of the funding announcement refreshes user profiles to surface a core feed of saved images at the forefront, closer in concept to a Tumblr blog than a folder structure. Cosmos is also beginning to publish trend reports derived from search and cluster data, seeded initially through its newsletter, as it moves to surface the niche aesthetic communities forming on the platform.