News for hackers. Stuff that matters.
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by 0cf8612b2e1e on Friday July 31, 2026 @10:15AM (#49123428)
Somewhat relatedly, how do the economics for Huggingface work? They must be hosting petabytes of models and datasets by now. I have downloaded quite a few “just in case”, only to replace them with the later iteration months later.

Does the file hosting actually cost peanuts when you do it yourself and the cloud has shattered my understanding of what it actually costs to deliver so much data?

Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by wongarsu on Friday July 31, 2026 @11:30AM (#49124378)
File hosting is pretty cheap. Egress traffic is about $90/TB in the cloud, but around $1/TB in the real world. Storage is in the realm of $5/TB/month after adding redundancy

At the scale of Huggingface, that still amounts to a lot load of money. Significantly less than if you did the same in AWS, but still a lot

That said, they do have a deal with AWS to make the data available in AWS ip space. Maybe they got some cheap hosting out of that too

Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by oceanplexian on Friday July 31, 2026 @01:04PM (#49125836)
It's less than $1/TB. If you have settlement-free peering it's like fractions of a penny at scale. Cloud provider DTO pricing is literally the biggest scam on Earth.
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by flyingpenguin on Friday July 31, 2026 @04:06PM (#49127968)
My company spends $8,000,000/month on storage and $4,000,000/month on data with aws... In the financial reports we do we round this number away. Make of that what you will
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by Bayart on Saturday August 01, 2026 @06:12AM (#49132962)
"Nobody ever got fired for buying AWS."
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by gandreani on Friday July 31, 2026 @01:35PM (#49126224)
What does DTO mean in this case? And settlement-free peering? I didn't know the price for S3 could be so cheap with AWS
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by lima on Friday July 31, 2026 @10:29AM (#49123633)
Bandwidth is really cheap when you run your own infra
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by himata4113 on Friday July 31, 2026 @12:02PM (#49124817)
It's not only cheap, it's actually free since if you are not using your 95th percentile you are losing money.
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by Scoundreller on Friday July 31, 2026 @04:15PM (#49128053)
Some enterprising ISPs get into the hosting business because most of their traffic is ingress, so adding some egress revenue is effectively zero cost.
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by miyuru on Friday July 31, 2026 @11:24AM (#49124293)
when trying to download models using browser, the DL domain for me us.aws.cdn.hf.co which points to bunch of singapore AWS IPv4 for me, despite it saying US.
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by jvuygbbkuurx on Friday July 31, 2026 @10:16AM (#49123453)
Maybe download numbers are relatively low
Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by rupx on Friday July 31, 2026 @11:55AM (#49124699)
Download numbers are reported on the site, and they aren't low.

CDN costs are pennies compared to inference and training though, HuggingFace will just get another 100 million and be set.

Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by anon373839 on Friday July 31, 2026 @07:22PM (#49129612)
HF uses chunk-level deduplication that greatly reduces the amount of physical storage needed for a given amount of logical storage.

https://huggingface.co/docs/xet/en/deduplication

Re:DeepSeek V4 Flash 0731 Intelligence, Performance and Pric... by Bayart on Saturday August 01, 2026 @06:11AM (#49132953)
Nowadays storage and bandwidth is cheap as long as you're not going through a hyperscaler and it doesn't require compute (ie. "dead" data). I'm working mostly on GCP so I'm not sure what are the rates now, but at least AWS were known for billing insane numbers on egress.

As long as you're smart on caching on the edges and deduplication/overlaying with your content (which structured data types like models, containers and repos typically are fit for) you can get remarkably far for less than you'd think.

The economics of serving models seem to be way more dodgy.