AI training-data specialist Snorkel AI announced on September 22, 2026 that it has closed a $350 million Series E co-led by Insight Partners and S32, valuing the San Francisco-based company at $3.5 billion. The round triples Snorkel's previous valuation and lands amid a scramble by frontier labs and enterprises for higher-quality specialist data.
Insight Partners And S32 Anchor A 16-Investor Round
Existing backer Addition returned alongside newcomers March Capital, Blumberg Capital, Allegis Capital, Frontline, Standard, Third Point Ventures, Greylock, Lightspeed, GV, Factory, Prosperity7, Walden Catalyst and Wells Fargo. Snorkel says annualized revenue has jumped 17x year over year, with its expert-in-the-loop Data-as-a-Service offering approaching $350 million in run-rate less than 12 months after launch.
Building The 'Data Factory' For Frontier AI
Founded in 2019 by researchers from the Stanford Snorkel Project, the company now serves six of the world's ten largest banks, top U.S. intelligence agencies, and multiple leading foundation-model labs. Co-founder and CEO Alex Ratner said the raise will accelerate investment in vertical and enterprise AI, extend research into new domains and modalities, and deepen the company's open research program.
"The teams pushing the frontier want a research data partner who pioneers the science of data development," Ratner said in the announcement.
A Bet On Post-Training And Agentic Data
Snorkel's Series E follows a wave of similarly aggressive rounds for AI-data specialists as reinforcement learning, agent evaluation and enterprise fine-tuning eat a growing share of frontier-lab budgets. The company's pitch, mirrored across recent partnerships in defense, finance and healthcare, is that programmatic labeling and expert curation now determine model quality more than raw pretraining compute.
For a broader look at how AI data budgets are reshaping venture activity this week, see coverage of Tekever's $580M Series D, D-Robotics' $400M Series C, and Anthropic's own agentic R&D push.
Reporting based on coverage from PR Newswire, TechCrunch, SiliconAngle and Unite.AI.