Snorkel AI Raises $350M at $3.5B Valuation to Scale Frontier AI Data Factory

Snorkel AI has raised $350 million at a $3.5 billion valuation in a new funding round aimed at scaling its data infrastructure for frontier and agentic artificial intelligence systems.

The San Francisco-based AI data company announced the round on September 22, 2026, saying the investment will expand its agentic data factory, which supplies the datasets, environments and evaluation systems used by advanced AI labs and enterprises.

The round was co-led by Insight Partners and S32, with significant participation from existing investor Addition. New investors in the round include March Capital, Blumberg Capital, Allegis Capital, Frontline, Standard and Third Point Ventures.

Existing investors Greylock, Lightspeed, GV, Factory, Prosperity7, Walden Catalyst and Wells Fargo also participated.

The new financing comes as AI development increasingly shifts from large-scale basic data labeling toward more complex datasets and environments designed for sophisticated models and AI agents.

From Data 1.0 to Data 2.0

Snorkel AI describes the changing AI data landscape as a transition from "Data 1.0" to "Data 2.0."

In the earlier generation of AI development, data preparation often involved relatively straightforward labeling tasks. The primary challenge was producing large volumes of labeled examples, making the process heavily dependent on scaling human labor.

Frontier AI systems require a different type of data.

According to Snorkel, modern agentic systems increasingly need expert-designed tasks, environments and evaluation rubrics that can take highly qualified people hours or even days to construct.

The company argues that developing these resources is no longer simply a data-labeling exercise. It increasingly resembles research work, where the complexity and quality of the resulting data can directly affect the performance of AI systems.

This shift forms the basis of Snorkel's agentic data factory.

Rather than focusing solely on producing large quantities of labeled information, the company is building infrastructure around the development of complex datasets, benchmarks, evaluations and environments for frontier AI.

Building Data for Frontier AI Labs

Snorkel AI works with leading AI labs and enterprises on the data used for frontier model training and evaluation.

The company launched its expert Data-as-a-Service offering in September 2025, and says it has grown rapidly since then.

Its work builds on research originating from Stanford AI Lab.

Snorkel says the company emerged from the Stanford AI Lab nearly a decade ago, with its founding team contributing to the development of data-centric AI.

The company's research foundation remains an important part of its current strategy.

According to Snorkel, its research work spans more than 250 peer-reviewed papers that have received more than 25,000 citations.

That research background is being incorporated into the company's commercial approach to AI data development.

Human Expertise Meets AI

A central element of Snorkel's model is combining human expertise with AI-driven systems.

The company describes its expert-agentic environments as creating a feedback loop between human expertise and artificial intelligence.

Highly specialized humans can help design complex tasks, environments and evaluation criteria, while AI systems can then operate within those environments.

This approach is particularly relevant as AI models become capable of performing longer and more complicated sequences of actions.

For companies developing frontier models, evaluating those systems requires more than checking whether an AI can generate a correct answer to an isolated question.

Models and agents increasingly need to be assessed on their ability to complete complex workflows, operate across environments and produce reliable outcomes.

Snorkel's data factory is designed to support those requirements.

Funding to Expand the Agentic Data Factory

The $350 million financing will primarily be used to increase the capacity of Snorkel's agentic data factory.

The company says demand for its data development capabilities is increasing as AI labs and enterprises work on more advanced systems.

Snorkel plans to expand its capacity to meet that demand while accelerating investment in vertical and enterprise AI.

The company also intends to extend its underlying research and technology into new domains and modalities.

Healthcare, law and software engineering are among the high-value industries highlighted by Insight Partners as areas where increasingly complex AI systems can benefit from research-grade data and evaluation infrastructure.

Snorkel also plans to increase its investment in open research.

The company specifically pointed to its Open Benchmarks Grants initiative as one of the programs it intends to build upon.

Investors Back AI Data Infrastructure

The new round brings together a combination of new and existing investors.

Insight Partners and S32 co-led the financing, while Addition made a significant follow-on investment.

March Capital, Blumberg Capital, Allegis Capital, Frontline, Standard and Third Point Ventures joined the company as new investors.

Existing investors including Greylock, Lightspeed, GV, Factory, Prosperity7, Walden Catalyst and Wells Fargo also participated.

The financing values Snorkel AI at $3.5 billion, according to the company's announcement.

Insight Partners Managing Director Lonne Jaffe said Snorkel's research-oriented approach to AI data, environments and measurement is becoming increasingly important for developing capable and reliable AI systems.

S32 CEO and General Partner Andy Harrison described Snorkel's expert-agentic environments as a way to connect human expertise and AI while producing data for model development.

A Different Layer of the AI Stack

Much of the attention around frontier AI has focused on model developers, computing infrastructure and applications.

Snorkel is focused on another part of the stack: the data and environments required to train and evaluate those systems.

As models become more capable, the complexity of the data required to measure and improve them also increases.

For agentic AI in particular, developers need to understand not only whether an agent can produce a correct response, but whether it can execute a sequence of tasks successfully within a defined environment.

That creates demand for specialized benchmarks, task environments and evaluation systems.

Snorkel's strategy is built around providing those resources to AI labs and enterprise customers.

Expanding Into New Domains and Modalities

The new capital will also allow Snorkel to extend its technology into additional domains and modalities.

The company has not specified every new area it plans to enter, but its investors highlighted industries such as healthcare, law and software engineering as areas where complex AI workflows could create demand for specialized data and evaluation.

The expansion also reflects the broader evolution of AI systems from general-purpose language models toward models and agents capable of working within specialized professional environments.

Different industries require different knowledge, workflows and standards for determining whether an AI system has performed correctly.

Developing those environments and evaluation systems requires domain expertise alongside technical infrastructure.

Continuing the Research Foundation

Snorkel's commercial strategy remains closely tied to its academic and research roots.

The company was founded out of the Stanford AI Lab in 2019 and continues to position research-driven data development as a core part of its offering.

Its body of published research, which the company says includes more than 250 peer-reviewed papers and more than 25,000 citations, provides the foundation for its approach to data-centric AI.

The company also intends to continue supporting open research through programs such as Open Benchmarks Grants.

This combination of commercial data services and open research is part of Snorkel's broader effort to develop infrastructure for the next generation of AI systems.

The Next Phase for Snorkel AI

With $350 million in new funding and a $3.5 billion valuation, Snorkel AI is preparing to expand the scale of its agentic data factory.

The company plans to increase capacity, invest further in enterprise and vertical AI, expand into new domains and modalities, and deepen its research efforts.

Its strategy reflects a broader change in the AI development process.

As frontier models move toward increasingly complex reasoning and agentic workflows, the data required to train and evaluate those systems is also becoming more specialized.

Snorkel AI is positioning itself around that emerging requirement, providing datasets, benchmarks, evaluation systems and custom environments designed for advanced AI development.

The latest funding gives the company additional resources to scale that infrastructure as AI labs and enterprises continue building increasingly sophisticated systems.