In recent years, creative marketplaces and platforms have realized that they are sitting on a treasure trove of data that they can use to develop AI models themselves, or license it to other AI labs to turn it into a revenue stream.
Wirestock, which previously helped photographers distribute and sell their work on stock photography services like Shutterstock, has chosen the latter path.
The company pivoted to being a data provider in 2023 and now provides datasets of images, videos, design assets, games and 3D content to AI Labs. Wirestock says its platform has more than 700,000 registered artists and designers who complete a variety of data-gathering tasks, similar to freelancers on platforms like Fiverr.
Mikael Khachatryan, the company’s co-founder and CEO, said Wirestock was transparent about its transition and allowed artists to opt out of the data-supplying business (in 2022, the platform had more than 100,000 photographers). Khachatryan declined to say how many people had switched to an AI data provider, saying only that “the vast majority” had done so.
“Initially, many of our deals were just selling off-the-shelf stuff, like existing libraries. But then that turned into a lot of custom requests for content and data, which created new opportunities for creators, and the platform took off quickly,” he said.
The company announced Thursday that it has raised $23 million in Series A funding to build a new data supply business. The round was led by Nava Ventures with participation from SBVP (co-founded by Sheryl Sandberg), Formula VC, and I2BF Ventures.
Khachatryan said Wirestock currently provides multimodal data to six of the largest basic model manufacturers, but he declined to name them. He said the company currently has annual run-rate revenue of $40 million and has paid out $15 million to contributors to date.
As part of the transition, the startup had to retrain some of its team to deeply annotate and label data to be useful in its AI lab. We also needed to build our sales and enterprise teams to pitch to hyperscalers and find ways to acquire more creative assets in areas like 3D modeling.
Wirestock currently uses email marketing and referral programs to acquire new contributors. Photographers, videographers and illustrators can apply for data on the website, but must complete a free task as a quality check before being accepted. The company said it evaluates all work on the platform using a combination of AI and human reviews.
As AI labs continue to refine their models, data feeding services are currently in high demand. Companies like Surge, Scale AI, and Mercor have built businesses worth tens of billions of dollars seemingly overnight on the back of demand for different kinds of datasets, and a series of startups like Micro1, Human Archive, and Human Native AI are also collaborating with top AI model makers.
Wirestock wants to focus on providing data for models that support creative use cases, such as image and video generation. The company is also exploring other avenues, such as audio and music.
Freddie Martignetti, founder of Nava Ventures, said his fund had been looking for startups that were innovating in sourcing and refining data before learning about Wirestock.
“I think Wirestock has a deep understanding of what fundamental models and hyperscalers need in terms of multimodal data to start creating more human-like systems. The basis of our thesis was that multimodal data will become increasingly important for models to complete real-world tasks, not just create images and videos,” Martinetti told TechCrunch.
Wirestock currently employs 60 people and plans to use the new funding to hire for research, engineering and product roles. We also build enterprise software for AI labs to collaborate on datasets.
This funding round brings the startup’s total funding to approximately $26 million.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
