Skip to content
View original post on X: ModalOfficial
Published
22/100

Snorkel AI runs millions of monthly agent evals on Modal, cutting sandbox costs 20%

AIModal says Snorkel AI uses its Sandboxes and Inference to run millions of agent evaluations each month, covering coding, infrastructure engineering and data science. Snorkel serves open-weight models and mixes CPU and GPU Sandboxes at scale, cutting its sandbox bill by 20%.
ModalVerified on XOfficial
@modal
View on X

Frontier AI data lab @SnorkelAI uses Modal Sandboxes and Modal Inference to power millions of agent evals each month.

Snorkel uses evals to test agents on rigorous and diverse benchmarks ranging from coding to infrastructure eng through data science.

Modal lets them flexibly serve open-weight models and use a mix of CPU and GPU Sandboxes at scale, cutting their sandbox bill by 20%.

Source details

Modal · x.comPublished · added here

Original title: Frontier AI data lab @SnorkelAI uses Modal Sandboxes and Modal Inference to power millions of agent evals each month.

Topics

AI agentsDeployment & engineeringProduct updates