# How Coactive uses Anyscale

**75%**: “an overall 75% reduction in GPU compute cost”

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

- Company: [Coactive](https://www.usedby.ai/companies/coactive.md)
- Tool: [Anyscale](https://www.usedby.ai/tools/anyscale.md)
- Teams: MLOps, Applied AI engineers, Platform team

## What the story says

Coactive AI runs its Multimodal AI Platform on Anyscale's Ray-based managed compute, deployed in its own Kubernetes clusters on AWS and Azure. It uses Ray Serve for model serving and large-scale processing of image and video data, with fractional GPU allocation and autoscaling.

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

> Anyscale delivered major cost gains through fractional GPU allocations, allowing Coactive to pack multiple model replicas onto a single GPU and directly reduce the number of GPUs needed for the same workload.

- **4x**: “an overall 4x cheaper per-image processing”
- **1 day**: “1 day to deploy new multimodal model endpoints, down from 1+ week”
- **25%**: “service definitions shrank to roughly 25% of their previous code size”

> One of our applied AI engineers said, ‘we should use this model,’ and the next day it was running in production. Before Anyscale, that would’ve taken a week or more.
>
> Ross Morrow, Principal Engineer

## What usedby checked

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

- Checked: the figure 75% is printed word for word on the page, near the name of Coactive.
- Checked: the passage quoted above is copied word for word from the page, near the name of Coactive.
- Checked: Coactive 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/coactive-ai-anyscale · How we check: https://www.usedby.ai/methodology