Snorkel AI raises $350 million Series E to scale its data factory for AI labs

Snorkel AI raises $350 million Series E to scale its data factory for AI labs

22 September 2026 — Snorkel AI, based in San Francisco, announced on 22 September 2026 that it has raised $350 million in a Series E round at a valuation it states as $3.5 billion. The round was co-led by Insight Partners and S32.

The round

• Amount and stage: $350 million, Series E, as stated in the company's own release.
• Lead investors: Insight Partners and S32.
• Other investors named by the company: Addition, with what the release calls significant participation; new investors March Capital, Blumberg Capital, Allegis Capital, Frontline, Standard and Third Point Ventures; and existing investors Greylock, Lightspeed, GV, Factory, Prosperity7, Walden Catalyst and Wells Fargo.
• Use of the money: to grow the capacity of what Snorkel calls its "agentic data factory", which produces training data and test environments; to invest further in AI for specific industries and enterprises; and to extend its research into new domains and data types.

What the company does

Snorkel AI began as a platform for labelling and curating data programmatically. It now describes itself as a frontier AI data lab: it supplies the data and the simulated environments that AI labs use to train and evaluate models, and it also builds specialised AI agents for enterprises. TechCrunch, reporting the same day, said the new valuation is about three times the company's previous one.

Why it matters for buyers

Demand for high-quality training data and evaluation environments has grown as model makers move from text answers to agents that take actions. For companies building their own AI systems, this round signals that specialist data suppliers are scaling up, which can widen the choice of providers for evaluation sets, red-teaming data and domain-specific training material.

What to check

• Enterprise buyers evaluating Snorkel's agent work should ask which parts are product and which are services, and how data used in their projects is kept separate from work done for AI labs.
• Teams comparing data vendors should ask for documented quality checks, the source of human-labelled data, and the licence terms that come with delivered datasets.

Sources

• Snorkel AI: Snorkel AI raises $350M to scale the data factory for frontier AI — 22 September 2026
• TechCrunch: Snorkel AI triples valuation to $3.5B as demand for AI training data booms — 22 September 2026