AI companies are spending more to build powerful models, but the data behind those systems is becoming a bigger business too. On September 22, 2026, Snorkel AI announced that it had raised $350 million at a $3.5 billion valuation. The company is seeing strong demand for advanced AI data. This includes expert-built training data, evaluation tasks, and simulated environments.
The Snorkel AI valuation has climbed from $1.3 billion in May 2025 to $3.5 billion in the latest round. That is roughly a 2.7-fold increase in 16 months. Insight Partners and S32 led the Series E round, with new and existing investors taking part.
Snorkel has also reported rapid business growth. Its annualized revenue run-rate has passed $375 million, up more than 18 times in less than a year. Much of that growth comes from its Data-as-a-Service business, which launched in September 2025.
Snorkel started with software for data labeling. Its business has since moved toward finished datasets, reinforcement-learning environments, and evaluation systems.
The difference becomes clearer with a coding example. A simple task asks a reviewer to mark whether a code sample is correct. A more advanced task starts with a realistic software problem. Experts then define the goal, create scoring rules, test the model, and review the result.
This is why AI training data is becoming more specialized. The value is shifting from the number of labels to the difficulty, quality, and purpose of the work.
“The teams pushing the frontier want a research data partner who pioneers the science of data development,” said Alex Ratner, co-founder and CEO of Snorkel AI. “That’s what Snorkel was built to be: the frontier lab for agentic data, combining human excellence with over a decade of research and technology.”
Where the data market is moving
- From labels to challenges: AI data teams are building tasks that test coding, reasoning, and decision-making.
- From checking to designing: Experts help create the task and grading system. AI tools then handle parts of data generation and quality checks.
- From static data to environments: Advanced models need the right environment to perform tasks. They also need to produce measurable results.
- From training to testing: Companies need data that shows whether a model performs well after training.
The Snorkel AI $3.5 billion valuation reflects a wider shift in the AI market. Investors are paying more attention to companies that provide the data and systems needed for advanced AI. These AI tools help businesses build and evaluate their models.
Snorkel says it works with AI labs, hyperscalers, enterprises, and the U.S. government. Its specialist network covers coding, law, and medicine, with coding among its biggest areas of demand.
The market is also seeing strong interest in specialized data businesses. Meta’s $14.3 billion deal for a 49% stake in Scale AI in 2025 highlighted the growing strategic value of AI data. For Snorkel, this demand has translated into rapid growth. Its Data-as-a-Service business launched in September 2025. Since then, its annualized revenue has surpassed $375 million.
Snorkel plans to use the new funding to expand its data factory and hire more researchers and engineers. The company also plans to grow its enterprise and government operations. It will enter new industries and work with new types of data. The company also plans to invest in AI research and model evaluations.
Author’s Note: The New York Press Release shows how the AI data market is moving beyond basic labeling. In my view, Snorkel AI’s latest funding is notable because its growth is tied to more complex data, expert-led tasks, and AI evaluation. That shift gives the $3.5 billion valuation more context and shows why specialized data is becoming a bigger part of AI development.
Leave a comment