Snorkel AI raised $350 million at a $3.5 billion valuation to fund its shift from data labeling software to providing finished training datasets and simulated environments. The round, led by Insight Partners and S32, values the seven-year-old company at nearly triple the price set during its Series D funding 17 months ago. Existing investors including Addition, Lightspeed, Greylock, GV, and Wells Fargo participated in the deal. Snorkel now generates an annualised revenue run-rate of $375 million, an eighteen-fold increase over the last year driven by demand for high-quality AI training data. Unlike competitors that act as marketplaces for human experts, Snorkel uses a hybrid model combining its software with subject matter specialists to generate synthetic data alongside human input. This approach allows the firm to sell reinforcement learning environments and complete datasets while keeping payments to human experts within cost of goods sold rather than gross revenue figures. The company launched commercially in 2019 after four years of research by co-founder and CEO Alex Ratner at a Stanford AI lab.
The valuation surge reflects a broader market trend where data companies positioning themselves as AI labs are seeing rapid growth. Mercor has reached $2 billion in gross annualised revenue, Handshake hit $1 billion earlier this year, and Micro1 scaled to $500 million. Because these firms pay 60% to 70% of their top-line income directly to domain specialists, their actual net revenue is substantially lower than headline figures. Snorkel’s model differs by selling finished products rather than pure labour, which impacts how its financial performance is reported. This distinction matters for investors assessing the true profitability of the data sector.
- Revenue run-rate stands at $375 million
- Valuation is nearly three times the Series D price
- CEO Alex Ratner previously worked at Stanford
