Wirestock Lands $23M Series A to Pipe Human-Made Multimodal Data Into Six of the Top Foundation Labs
The 700,000-creator platform now runs at a $40M revenue clip and is becoming the back-end for the labs racing to push frontier models past pure web-scrape data.
Wirestock has closed a $23 million Series A led by Nava Ventures, with participation from SBVP — co-founded by Sheryl Sandberg — alongside Formula VC, I2BF Global Ventures, and existing backers. The round bankrolls a platform that has quietly become one of the most important upstream suppliers of human-made multimodal training data to the frontier AI labs.
Founded by CEO Mikayel Khachatryan, Wirestock connects more than 700,000 creators — photographers, videographers, motion designers, 3D artists, and musicians — to AI labs that need licensed, ethically sourced data for model training. The company says it now serves six of the largest foundation model makers, runs at $40 million in annual run-rate revenue, and has paid out $15 million to contributors to date. Headcount sits at around 60.
The business has shifted decisively from off-the-shelf stock to bespoke data orders. "Initially, a lot of our deals were just selling what we had off the shelf," Khachatryan told TechCrunch. "But then it turned into a lot of custom requests for content and data." Nava's Freddie Martignetti framed the thesis bluntly: "Multimodal data will be increasingly important, not just to create images or videos, but for models to complete real-world tasks."
The funding will go into expanding the creator platform with new earning pathways and tooling for working artists, plus deeper investment in video, audio, and gaming data. The raise comes as OpenAI, Anthropic, and Google all increase spending on licensed training data after a wave of copyright suits — including the New York Times case — pushed labs to formalize human-sourced data pipelines instead of relying on open-web scraping.
Comments
Share your thoughts. Be kind.
Loading comments…
