Snorkel AI, a startup that helps AI research groups and corporate teams produce training datasets and simulated environments, has closed a $350 million Series E round that values the company at $3.5 billion.
The round was led by Insight Partners and S32, and existing investors including Addition, Lightspeed, Greylock, GV, and Wells Fargo also participated. The new valuation is nearly three times the $1.3 billion mark Snorkel achieved 17 months earlier when it raised $100 million in a Series D.
Shift from labeling software to data-as-a-service
Originally Snorkel offered software to automate data labeling. Last year the company shifted its go-to-market approach to providing finished datasets and simulated environments as a data-as-a-service product. Rather than operating solely as a marketplace for human experts, Snorkel uses a hybrid model that combines its software and models to generate synthetic data alongside contributions from subject-matter experts.
Snorkel says this repositioning and surging demand for high-quality training data have driven rapid growth: the company’s current annualized revenue run-rate stands at $375 million, an 18-fold increase over the past 12 months.
Industry context and cost structure
Other firms that position themselves as AI data labs have reported similar headline growth. Mercor’s gross annualized revenue has grown to $2 billion, Handshake reached $1 billion earlier this year, and TechCrunch reported Micro1 scaling to $500 million. It’s important to note that many of these companies pay roughly 60–70% of top-line revenue directly to the domain specialists performing the work, so their net revenue after payouts is substantially lower than the gross figures.
Snorkel distinguishes its model by selling reinforcement-learning environments and complete datasets rather than directly selling human labor. According to the company, payments to human experts are recorded in cost of goods sold (COGS), not added to the headline annualized revenue figures.
Origins and outlook
Snorkel launched commercially in 2019 after four years of research by co-founder and CEO Alex Ratner and his team at a Stanford AI lab. The company’s recent funding round and the reported surge in run-rate revenue underscore ongoing strong demand from AI labs for high-end training data and simulated environments.
