Key facts
- Snorkel AI raised $350 million in a funding round led by Insight Partners and S32.
- The company's valuation increased to $3.5 billion, nearly tripling from its May 2025 valuation of $1.3 billion.
- Snorkel AI's annualized revenue run-rate has surpassed $350 million, up from approximately $20 million a year prior.
- The company shifted from selling software to supplying finished datasets and reinforcement-learning environments.
- Snorkel AI uses a platform that combines human experts with AI models to create and vet data.
- The company works with frontier AI labs, hyperscalers, enterprises, and the US federal government.
Data startup Snorkel AI has secured $350 million in new funding, achieving a valuation of $3.5 billion, according to CEO Alex Ratner. The investment was driven by a surge in demand for complex training data and simulated environments from leading AI labs. This latest funding round nearly triples the company's valuation from its previous $1.3 billion valuation when it raised $100 million in May 2025.
Snorkel AI, based in San Francisco, reported that its annualized revenue run-rate has now surpassed $350 million, a significant increase from approximately $20 million a year earlier. This growth is attributed to its data-as-a-service business launched in September 2025.
Founded in 2019 by researchers from Stanford University, Snorkel AI initially focused on software but has since pivoted to providing finished datasets and reinforcement-learning (RL) environments. Ratner explained that AI developers are increasingly seeking more sophisticated data for training advanced systems, moving beyond simpler labeling tasks.
The company's platform, described as an "agentic data development platform," combines human expertise with AI models to create and validate data. Human specialists devise scenarios and grading criteria, while AI automates quality assurance. Snorkel AI partners with AI labs to develop high-quality datasets, leveraging a network of tens of thousands of specialists in fields like coding, law, and medicine. The company sells the generated data products rather than charging for human labor, which allows for competitive expert compensation while maintaining profit margins.
Snorkel AI serves frontier AI labs, hyperscalers, enterprises, and the U.S. federal government, with coding data being a significant area of demand. The venture capital market has seen substantial investment in startups providing human-annotated training data to AI labs, a trend highlighted by Meta's acquisition of a stake in Scale AI in June 2025. Rivals such as Mercor and Surge AI have also attracted investor interest.
The newly raised capital will be used to hire researchers and engineers, expand enterprise and government operations, support third-party AI model evaluations, and enter new industry verticals and data modalities. The company anticipates reaching profitability this year.