Data-labeling and evaluation startup Snorkel AI has raised $350 million in a round led by Insight Partners and S32, with participation from Addition, Greylock, and Wells Fargo, valuing the company at $3.5 billion. CEO Alex Ratner confirmed the figures to Reuters.
That valuation is nearly triple the $1.3 billion mark reported in May 2025. Coverage from AI Weekly and AI Chat Daily says annual recurring revenue jumped from about $20 million to as high as $350–$375 million in a year — on the order of a 17–18x increase.
The driver is demand from frontier labs for complex training data and simulated reinforcement-learning environments, not just classic annotation. As models get used as agents, labs are paying for curated tasks, expert feedback, and environments that look more like software testbeds than spreadsheets of labels.
Snorkel’s raise is a reminder that the AI stack’s bottleneck has shifted. Chips and base models still dominate headlines, but the companies that manufacture high-quality training signal are now raising at multi-billion-dollar valuations of their own.
Key takeaway. A $3.5 billion valuation on $350 million of new capital shows how fast “training data and RL environments” has become a core industry, not a side vendor category.
Photo: Unsplash. Sources: Reuters exclusive; AI Weekly; AI Chat Daily (Sept. 22–23, 2026).
