CoreWeave AI Object Storage read throughput
Company: CoreWeave
The claim, verbatim
CoreWeave AI Object Storage served 110 to 120 GB/s throughput for reading datasets
Source (primary)
Video Captioning at Scale: 600 TB in 95 Minutes | CoreWeave Blog - CoreWeave (-, news_article)
View cached copy (2026-09-01)Live source ↗
Quote: “CoreWeave AI Object Storage served 110 to 120 GB/s across roughly 46 CPU nodes”
How we checked this
Checked on September 18, 2026. The cited source supports every part of this claim.
The post states that storage served 110 to 120 GB/s, across roughly 46 CPU nodes, while the full 600 TB dataset was being captioned.
Confirmed in the source:
- CoreWeave AI Object Storage served 110 to 120 GB/s
- This throughput was for reading the 600 TB dataset during the captioning run, without caching
What we did: Read our cached copy of the publisher (https://www.coreweave.com/blog/video-captioning-at-scale-600-tb-in-95-minutes-with-anyscale-on-coreweave) in full (16,024 characters, retrieved September 1, 2026) and checked each assertion in the claim against it.
Additional evidence
confirms Video Captioning at Scale: 600 TB in 95 Minutes | CoreWeave Blog - CoreWeave
Quote: “CoreWeave AI Object Storage served 110 to 120 GB/s across roughly 46 CPU nodes”