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

Are OpenAI and Anthropic losing money on inference?

martinalderson.com

Post reply on HN
Print view View original

161–170 of 495 posts

Re: Are OpenAI and Anthropic losing money on inference?

#161

Post by GaggiX »

noodletheworld wrote: post #145

Huh. I feel oddly skeptical about this article; I can't specifically argue the numbers, since I have no idea, but... there are some decent open source models; they're not state of the art, but if inference is this cheap then why aren't there multiple API providers offering models at dirt cheap prices? The only cheap-ass providers I've seen only run tiny models. Where's my cheap deepseek-R1? Surely if its this cheap,…

https://openrouter.ai/deepseek/deepseek-chat-v3.1

They are dirt cheap. Same model architecture for the comparison: $0.30/M $1.00/M. Or even $0.20-$0.80 from another provider.

Re: Are OpenAI and Anthropic losing money on inference?

#162

Post by Filligree »

nativeit wrote: post #147
EcommerceFlow wrote: post #94

Earlier quoted context omitted.

A lifetime of using Google and 4 years of using LLMs.

…is a great counter-example of a “source”. It’s not like the product at-hand is relevant to data analysis or anything, amirite?

Sometimes a statement is just too obvious to need extensive sourcing, and this is one of those times.

Gemini doesn’t always find very much better results, but it usually does. It beggars belief to claim that it doesn’t also understand the query much better than Rankbrain et al.

Re: Are OpenAI and Anthropic losing money on inference?

#163

Post by martinald »

scrollaway wrote: post #76
martinald wrote: post #13

Earlier quoted context omitted.

(Author here). Yes I am aware of that and did mention it. However - what I wanted to push back in this article was that claude code was completely unsustainable and therefore a flash in the pan and devs aren't at risk (I know you are not saying this). The models as is are still hugely useful, even if no further training was done.

> claude code was completely unsustainable and therefore a flash in the pan and devs aren't at risk How can you possibly say this if you know anything about the evolution of costs in the past year? Inference costs are going down constantly, and as models get better they make less mistakes which means less cycles = less inference to actually subsidize. This is without even looking at potential fundamental improvements…

I totally agree with you! I have heard others saying this though. But I don't think it's true.

Re: Are OpenAI and Anthropic losing money on inference?

#164

Post by hirako2000 »

noodletheworld wrote: post #145

Huh. I feel oddly skeptical about this article; I can't specifically argue the numbers, since I have no idea, but... there are some decent open source models; they're not state of the art, but if inference is this cheap then why aren't there multiple API providers offering models at dirt cheap prices? The only cheap-ass providers I've seen only run tiny models. Where's my cheap deepseek-R1? Surely if its this cheap,…

Imo the article is totally off the mark since it assumes users on average do not go over th 1M tokens per day.

Afaik openai doesn't enforce a daily quota even on the $20 plans unless the platform is under pressure.

Since I often consume 20M token per day, one can assume many would use far more than the 1M tokens assumed in the article's calculations.

Re: Are OpenAI and Anthropic losing money on inference?

#165

Post by gitremote »

re-thc wrote: post #118
Topfi wrote: post #114

Earlier quoted context omitted.

That doesn't seem compatible with what he stated more recently: > We're profitable on inference. If we didn't pay for training, we'd be a very profitable company. Source: https://www.axios.com/2025/08/15/sam-altman-gpt5-launch-chat... His possible incentives and the fact OpenAI isn't a public company simply make it hard for us to gauge which of these statements is closer to the truth.

> That doesn't seem compatible with what he stated more recently: Profitable on inference doesn't mean they aren't losing money on pro plans. What's not compatible? The API requests are likely making more money.

Yes, API pricing is usage based, but ChatGPT Pro pricing is a flat rate for a time period.

The question is then whether SaaS companies paying for GPT API pricing are profitable if they charge their users a flat rate for a time period. If their users trigger inference too much, they would also lose money.

Re: Are OpenAI and Anthropic losing money on inference?

#166

Post by jsnell »

noodletheworld wrote: post #145

Huh. I feel oddly skeptical about this article; I can't specifically argue the numbers, since I have no idea, but... there are some decent open source models; they're not state of the art, but if inference is this cheap then why aren't there multiple API providers offering models at dirt cheap prices? The only cheap-ass providers I've seen only run tiny models. Where's my cheap deepseek-R1? Surely if its this cheap,…

> why aren't there multiple API providers offering models at dirt cheap prices?

There are. Basically every provider's R1 prices are cheaper than estimated by this article.

https://artificialanalysis.ai/models/deepseek-r1/providers

Re: Are OpenAI and Anthropic losing money on inference?

#167

Post by martinald »

noodletheworld wrote: post #145

Huh. I feel oddly skeptical about this article; I can't specifically argue the numbers, since I have no idea, but... there are some decent open source models; they're not state of the art, but if inference is this cheap then why aren't there multiple API providers offering models at dirt cheap prices? The only cheap-ass providers I've seen only run tiny models. Where's my cheap deepseek-R1? Surely if its this cheap,…

There are, I screenshotted DeepInfra in the article, but there are a lot more https://openrouter.ai/deepseek/deepseek-r1-0528

Re: Are OpenAI and Anthropic losing money on inference?

#168

Post by hirako2000 »

gitremote wrote: post #45

These numbers are off. > $20/month ChatGPT Pro user: Heavy daily usage but token-limited ChatGPT Pro is $200/month and Sam Altman already admitted that OpenAI is losing money from Pro subscriptions in January 2025: "insane thing: we are currently losing money on openai pro subscriptions! people use it much more than we expected." - Sam Altman, January 6, 2025 https://xcancel.com/sama/status/1876104315296968813

Trusting the man about costs would be even more misplaced than trusting an oil company's CEO about the environment.

Re: Are OpenAI and Anthropic losing money on inference?

#169

Post by vkou »

CjHuber wrote: post #148
Topfi wrote: post #114

Earlier quoted context omitted.

That doesn't seem compatible with what he stated more recently: > We're profitable on inference. If we didn't pay for training, we'd be a very profitable company. Source: https://www.axios.com/2025/08/15/sam-altman-gpt5-launch-chat... His possible incentives and the fact OpenAI isn't a public company simply make it hard for us to gauge which of these statements is closer to the truth.

Does anybody really think in this current time that what a CEO says has anything to do with reality and not just with hyping up ala elon recipe

Specifically, a connected CEO in post-law America.

This sort of thing used to be called fraud, but there's zero chance of criminal prosecution.

Post reply on HN