Re:The Unbearable Cheapness of Open Weight Models
by uejfiweun on Thursday June 25, 2026 @07:01PM
(#48680293)
What are you even talking about? Everyone knows that Anthropic is drastically subsidizing their plans. It's actually the exact opposite of what you're talking about. The costs are extremely high and the prices are actually what's being subsidized and cheap right now.
[ Parent
| on HN ]
Re:The Unbearable Cheapness of Open Weight Models
by Tuna-Fish on Thursday June 25, 2026 @07:26PM
(#48680515)
This is an example of common knowledge that is wrong. People look at their cash burn, assume that they spend this to subsidize inference, and get bonkers answers. Inference is not their largest expense.
Inference is cheap. Anthropic is only drastically subsidizing their plans if you count their training expenses as part of their costs.
Re:The Unbearable Cheapness of Open Weight Models
by therobots927 on Thursday June 25, 2026 @09:15PM
(#48681251)
Are you an anthropic insider or something? Because if you are you should delete this comment. If you aren’t then you don’t know what the hell you’re talking about.
Re:The Unbearable Cheapness of Open Weight Models
by xquce on Friday June 26, 2026 @02:08AM
(#48682908)
Surely the same can be said for the people saying the opposite?
Re:The Unbearable Cheapness of Open Weight Models
by therobots927 on Friday June 26, 2026 @07:29AM
(#48685382)
I didn’t make a claim. The parent explicitly said it was a misconception that inference is not profitable.
No one knows if it’s profitable or not so we’re left to speculate.
Re:The Unbearable Cheapness of Open Weight Models
by Bnjoroge on Friday June 26, 2026 @09:35AM
(#48686498)
You can make a fairly decent assumption by calculating the margin on serving glm 5.2, and adding say 30% extra costs and it still leaves a healthy margin
Re:The Unbearable Cheapness of Open Weight Models
by therobots927 on Friday June 26, 2026 @10:17AM
(#48686891)
Where are you getting 30% from
Re:The Unbearable Cheapness of Open Weight Models
by Bnjoroge on Friday June 26, 2026 @10:51AM
(#48687337)
It was a rough heuristic for how much more opus/5.5 would presumably cost if you extrapolate from glm5.2 prices. In any case input tokens for 5.4/4.6 are 70-100% more expensive, cached about 3-100%, and output tokens anywhere from 60%-240% more as per all their current api pricing. I highly doubt 5.4/4.6 are that much more expensive to serve given how cheap and commoditized inference has become, and how comparable they are perfomance wise.
Re:The Unbearable Cheapness of Open Weight Models
by rcxdude on Friday June 26, 2026 @09:09AM
(#48686194)
For one point, you can look at the costs of similarly sized open-source models from inference providers (which are only making money on the markup on the compute), and compare with anthropic's prices. There's a pretty big price difference there and it would be hard to believe that anthropic's models are that much more expensive to run than those models.
Re:The Unbearable Cheapness of Open Weight Models
by tomnipotent on Friday June 26, 2026 @05:55PM
(#48692477)
Those prices don't need to bake in the training cost, since it was eaten by someone else (whomever trained the open source models). Anthropic et al. need to price in the whole lifecycle.
Re:The Unbearable Cheapness of Open Weight Models
by rcxdude on Saturday June 27, 2026 @07:32AM
(#48697336)
Yes, but the point is specifically on whether inference in and of itself is profitable (i.e. whether the unit economics work out). They are still losing money overall, but they're not losing more money the more people use their product (quite the opposite: they need a lot of users at their fairly large per-token margins in order to justify their R&D spend, and the big question IMO is how strong the opposite side is: how much money do they need to spend on R&D to have a product that justifies such a high per-token premium?).
Re:The Unbearable Cheapness of Open Weight Models
by dathanb82 on Friday June 26, 2026 @01:27AM
(#48682654)
If "inference is cheap," why is OpenAI spending a ton getting Broadcom to design custom AI chips that make inference cheaper? Reports suggest their custom silicon isn't all that good for training, it's all to make inference more efficient. That shouldn't be necessary if inference is already quite cheap.
Re:The Unbearable Cheapness of Open Weight Models
by petra on Friday June 26, 2026 @04:50AM
(#48684112)
A large part of the market will be ad based. For that, having the lowest cost inference is useful.
Also for agent doing r&d, cheaper tokens allows doing more, which is always good.
Re:The Unbearable Cheapness of Open Weight Models
by topranks on Saturday June 27, 2026 @04:56AM
(#48696446)
But the training expense is part of their costs!
The question is can they just stop training at some point, fixing the models in time, and still have a useful product.
Re:The Unbearable Cheapness of Open Weight Models
by Bnjoroge on Friday June 26, 2026 @09:34AM
(#48686481)
I dont think that’s accurate. I mean look at how much more expensive frontier closed source models are vs something like glm 5.2 which is just about as good. Serving glm is really cheap, and high margin. Obviously no one knows, just how much their inference costs, but if we assume that opus/gpt are maybe 15-20% more parameters than glm 5.2, then it makes no sense for them to charge almost much much more than glm