Micro1, a four-year-old AI data startup, has expanded its gross annual run rate from $100 million to $500 million over the past eight months, putting its net annual run rate between $150 million and $200 million after retaining roughly 60% to 70% of revenue. While the startup lags competitors like Mercor ($2 billion gross annualized revenue) and Handshake ($1 billion), its rapid growth demonstrates strong demand for AI training data suppliers, with some researchers predicting that future AI spending on data could rival spending on compute.
The company is increasingly generating synthetic data without human involvement, such as automated descriptions of video content. Selling the same datasets to multiple clients has driven gross margins for "off-the-shelf" data as high as 80% to 90%, though this practice has sparked controversy over concerns that such data may help Chinese AI developers build competitive models. Founder Ali Ansari stated that unlike some competitors, Micro1 does not sell data to Chinese model makers.
Micro1 raised its Series A at a $500 million valuation last September, and sources indicate the startup may have recently raised another round at a significantly higher valuation. The company, which began as an AI recruiting startup before pivoting to data-labeling, is also building a robotics pre-training dataset by having hundreds of generalists record everyday object interactions in their homes.