Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

LLMs are cheap

snellman.net

Post reply on HN
Print view View original

21–30 of 319 posts

Re: LLMs are cheap

#21

Post by jsnell »

Etheryte wrote: post #7

> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…

Ok, I clearly should have made the wording more explict since this is the second comment I got in the same vein. I'm not saying you'd convert users to $1/month subscriptions. That would indeed be an absurd idea.

I'm saying that good-enough LLMs are so cheap that they could easily be monetized with ads, and it's not even close. If you look at other companies with similar sized consumer-facing services monetized with ads, their ARPU is far higher than $1.

A lot of people have this mental model of LLMs being so expensive that they can’t possibly be ad-supported, leaving subscriptions as the only consumer option. That might have been true two years ago, but I don't think it's true now.

Re: LLMs are cheap

#22

Post by andrew_lettuce »

Etheryte wrote: post #7

> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…

Of there 500M users a very small number are already paying, so it's not zero-to-one for all of them, but monetize more and take $10 a month to $100. It's unclear if this is easier or harder than what you presented, but both are hard.

Re: LLMs are cheap

#24

Post by pama »

xxbondsxx wrote: post #5

You can't compare an API that is profitable (search) to an API that is likely a loss-leader to grab market share (hosted LLM cloud models). Sure there might not be any analysis that proves that they subsidized, but you also don't have any evidence that they are profitable. All the data points we have today show that companies are spending an insane amount of capex on gaining AI dominance without the revenue to achiev…

Please read the DeepSeek analysis of their API service (linked in this article): they have 500% profit margin and they are cheaper than any of the US companies serving the same model. It is conceivable that the API service of OpenAI or Anthropic have much higher profit margins yet.

(GPUs are generally much more cost effective and energy efficient than CPU if the solution maps to both architectures. Anthropic certainly caches the KV-cache of their 24k token system prompt.)

Re: LLMs are cheap

#25

Post by dist-epoch »

Etheryte wrote: post #7

> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…

This is true only because people are so dumb.

Paying $1000 for an iPhone? Sure. $10 for a Starbucks? Sure. $1 per year for LLM? Now hold on, papa is not an oil oligarch...

Re: LLMs are cheap

#26

Post by exceptione »

theOGognf wrote: post #17

Some anecdotal data, but we recently estimated the cost of running a LLM at $WORK by looking at power usage over a bursty period of requests from our internal users and it was on the order of $10s/mil tokens. And we arent a big place, nor were our servers at max load, so I can see the cost being much lower at scale

This is only the power usage?

Re: LLMs are cheap

#27

Post by iwontberude »

falcor84 wrote: post #3
datadrivenangel wrote: post #2

low to moderate quality digital text work is now almost free! This is going to reshape large portions of our text based communication networks.

There's of course also the issue that an increasing fraction of web content reading is being done by AI agents. I wonder what the Pareto front here is.

No one has successfully rebutted that paper about stochastic collapse of AI models which happens when models train on their own output over time. It’s just a matter of time before we find out if it was right or not.

Re: LLMs are cheap

#28

Post by ETH_start »

Search is narrow, used occasionally to find external information. LLMs are the single most general-purpose tool in existence. If you're using them to their full potential, you end up relying on them across writing, planning, coding, summarizing, etc.

So even if the per-query or per-token cost is lower, the total consumption is vastly higher. For that reason, while it may not be a fair comparison, due to people looking at it from the perspective of personal economics, people will compare how much it costs to use each to its full potential, respectively.

Re: LLMs are cheap

#29

Post by chaz6 »

Etheryte wrote: post #7

> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…

I thought that services like these were run at a loss because the data that users provide is often worth more than the price of a subscription.

Re: LLMs are cheap

#30

Post by hackyhacky »

qoez wrote: post #4

So far. Give it a few years when the core players have spent their way to market dominance and regulation kicks in and you'll see the price hikes investors have been promised behind closed doors.

Or maybe they'll just use ads.

Whatever question you ask, the response will recommend a cool, refreshing Coca Cola soft drink.

Your AI coding project will automatically display ads collecting revenue for Anthropic, not for you.

Every tenth email sent by your AI agent will encourage the recipient to consider switching to Geico.

The opportunities are endless.

Post reply on HN