# How Multiply Labs uses Anyscale

**30 to 10 minutes**: “cluster spin-up with H100s on Nebius used for training was reduced from 30 to 10 minutes”

As published on [anyscale.com](https://www.anyscale.com/resources/case-study/multiply-labs). Captured by usedby on 2026-10-07.

- Company: [Multiply Labs](https://www.usedby.ai/companies/multiply-labs.md)
- Tool: [Anyscale](https://www.usedby.ai/tools/anyscale.md)
- Teams: Robotics R&D

## What the story says

Multiply Labs uses Anyscale-managed Ray clusters to run distributed training of physical AI and robotics models, including imitation learning and Vision-Language-Action approaches, for biomanufacturing robots. It deploys them in its own cloud accounts and across AWS, GCP and Nebius to get access to premium GPUs.

Summary written by usedby from the source page, in English. The figures are those of Anyscale and Multiply Labs, not ours.

> As a result, cluster spin-up with H100s on Nebius used for training was reduced from 30 to 10 minutes with similar results for other GPU types across cloud providers.

> Using Anyscale as our multi-cloud Ray cluster manager, we’ve made our training runs fully portable across hyperscalers and specialized cloud providers, maximizing GPU availability.
>
> Varun Bhatia, Staff Robotics Software Engineer

## What usedby checked

We compared the story with its live page on 2026-10-07.

- Checked: the figure 30 to 10 minutes is printed word for word on the page, near the name of Multiply Labs.
- Checked: the passage quoted above is copied word for word from the page, near the name of Multiply Labs.
- Checked: Multiply Labs uses Anyscale. Confirmed line. Latest check across sources: 2026-10-07.
- Not checked: the result itself. We quote it; we did not measure it.

---
Source: https://www.usedby.ai/case-studies/multiply-labs-anyscale · How we check: https://www.usedby.ai/methodology