No, it doesn't cost Anthropic $5k per Claude Code user • Hacker News Board

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#311

Post by dathinab » Tue, Mar 10, 2026, 5:59 PM UTC

ffsm8 wrote: post #130

codemog wrote: post #115

Earlier quoted context omitted.

[flagged]

Omg, I can't believe that's real I wanted to believe that you're essentially trolling, but no - that service exist. And not an upstart, there is coverage going back several years. Our societies are seriously fucked.

to be fair dogs watching TV isn't really new, people putting up some YT play list or similar is not that uncommon

kinda like Netflix and YT have "fireplace" streams or how LG TVs can be setup as "digital picture frames" when not "actively" used

but it being a dedicated service people pay money for is something new for me too

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#312

Post by re-thc » Tue, Mar 10, 2026, 6:17 PM UTC

atq2119 wrote: post #218

re-thc wrote: post #188

Earlier quoted context omitted.

> Both Amazon and Google serve Opus at roughly ~1/2 the speed of the chinese models We were responded about 10x not 0.5x. x86 vs arm64 could have different performance. The Chinese models could be optimized for different hardware so it could show massive differences.

These providers do not run models on CPUs, x86 vs. Arm is irrelevant.

They run Nvidia and Huawei for example. And mine was just an example.

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#313

Post by stusmall » Tue, Mar 10, 2026, 6:41 PM UTC

Verdex wrote: post #303

nr378 wrote: post #295

Earlier quoted context omitted.

Dario has made a specific cohort argument here. His numbers (from various interviews) are: you train a model in 2023 for $100M, deploy it, and it earns $200M over its lifetime. Meanwhile you train the 2024 model for $1B, which goes on to earn $2B. Each vintage returns 2x on its training cost. However, the GAAP P&L tells the opposite story. You book $200M revenue in the same year you spend $1B training the next model,…

That's an interesting idea. I'm curious, though, are there any other industries and/or companies that have tried to pull this sort of thing off? And what ultimately happened to them?

Enron had a system like this. They regularly worked on large, long term contracts that became profitable over years/decades. They wanted to push rewards forward so would estimate the total value of the contract and book the profit when it closed. Mark-to-market accounting wasn't unheard of the time but using it for assets without an active market was unique. Without the market to make against, the numbers were best guess projections.

The problem is everyone along the line is incentivized to be aggressive with estimate (commissions for sales are bigger, public financials looks better) and discouraged from correcting the estimates when they go wrong.

Estimating multi-year returns on frontier models looks harder than estimating returns on oil and gas projects in the 90s.

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#314

Post by Chamix » Tue, Mar 10, 2026, 6:41 PM UTC

johndough wrote: post #304

Chamix wrote: post #293

Earlier quoted context omitted.

What do you think labs are doing with the minimum 10TB memory in NvLink 72 systems that were publicly reported to all start coming online in November/December of last year? And why would this 1 TB -> 10 TB jump matter so much for Anthropic previously being wholly dependent on running Opus 4x on TPUs, if the models were 2-3T at 4bit and could fit in 8x B200 (1.5 TB = 3T param) widely deployed during the Opus 4 era? Yo…

Could you point at some more public info about active parameter count? You said: > and while an exact number is hard to compute, let me tell you, it is not 17B or anywhere in that particular OOM :) I can see ~100B, but that would near the same order of magnitude. I find ~1000B active parameters hard to believe.

Sorry if that was unclear, I did mean 100Bs as in the next order of magnitude. Even GPT-4 had ~220B active params, though the trend has been towards increased sparsification (lower activation:total ratio). GPT 4.5 is the only publicly facing model that approached 1T active parameters (an experiment to see if there was any value in the extreme inference cost of quadratically increasing compute cost with naïve-like attention). Nowadays you optimize your head size to your attention kernel arch and obtain performance principally through inference time scaling (generate more of tokens) and parallel consensus (gpt pro, gemini deep think etc), both of which favor faster, cheaper active heads.

4o and other H100 era models did indeed drop their activated heads far smaller than gpt-4 to the 10s just like current Hopper-Era Chinese open-source, but it went right back up again post-Blackwell with the 10x L2 bump (for kv cache) in congruence with nlogn attention mechanisms being refined. Similar story for Claude.

The fun speculation is wondering about the true size of Gemini 3's internals, given the petabyte+ world size of their homefield IronwoodV7 systems and Jim Keller's public penchant for envisioning extreme MoE-like diversification across hundreds of dedicated sub-models constructed by individual teams within DeepMind.

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#315

Post by yunwal » Tue, Mar 10, 2026, 6:41 PM UTC

Verdex wrote: post #303

nr378 wrote: post #295

Earlier quoted context omitted.

Dario has made a specific cohort argument here. His numbers (from various interviews) are: you train a model in 2023 for $100M, deploy it, and it earns $200M over its lifetime. Meanwhile you train the 2024 model for $1B, which goes on to earn $2B. Each vintage returns 2x on its training cost. However, the GAAP P&L tells the opposite story. You book $200M revenue in the same year you spend $1B training the next model,…

That's an interesting idea. I'm curious, though, are there any other industries and/or companies that have tried to pull this sort of thing off? And what ultimately happened to them?

The bar for "wildly unprofitable" has risen quite a bit since then, but Amazon basically pioneered this.

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#316

Post by aurareturn » Tue, Mar 10, 2026, 6:50 PM UTC

bandrami wrote: post #235

aurareturn wrote: post #172

Earlier quoted context omitted.

How are we so sure that deep inside the moon isn't made out of cheese?

I remember Enron. Hell, I remember the S&Ls. I've seen this movie too many times to not know how it ends.

I remember Google, Meta, Apple, Eli Lily, and other meteoric risen companies.

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#317

Post by bdangubic » Tue, Mar 10, 2026, 6:53 PM UTC

addandsubtract wrote: post #288

itintheory wrote: post #248

Earlier quoted context omitted.

Any references on this? I hear this argument a lot. In fact, in a talk on AI last week I heard someone say: "If you click the thumbs up button to rate a chat, the AI provider will use the contents for training, so our company's policy is never to click the thumbs up button" That seemed so farcical I had a hard time taking this person seriously. Enterprise plans must give some strong guarantees around data usage, righ…

This is why most businesses only have ChatGPT subscriptions. Plus their integration into existing Microsoft products and billing.

Trusting Microsoft seems like a right move /s

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#318

Post by tgrowazay » Tue, Mar 10, 2026, 7:37 PM UTC

kleton wrote: post #271

tcbrah wrote: post #219

Earlier quoted context omitted.

yeah the json token counts are super misleading. i run a bunch of claude agents for automation and like 85% of input tokens end up being cached reads -which cost 1/10th of the sticker price. so your $200k number is probably closer to $25-30k in real cost, and thats before you factor in that anthropics own infra is way cheaper than retail API pricing. the $5k forbes number was always nonsense but even the "corrected"…

I proxy all of my llm completion subscriptions. In a typical 7d span- model completions read write cached_read cache_write claude-opus-4-6 11000 16900000 5840000 1312000000 66120000

17M uncached reads (input) and 6M of uncached writes (output) is

  $5x17+$25x6=$235 for Opus 4.6

  $2x17+$12x6=$106 for Gemini 3 Pro

  $0.60x17+$3.6x6=$31.80 for Qwen3.5 397B-A17B via Huggingface API

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#319

Post by clbrmbr » Tue, Mar 10, 2026, 7:52 PM UTC

bakugo wrote: post #118

Weaver_zhu wrote: post #104

Earlier quoted context omitted.

Agree, but I guess the Opus 4.6 is 10x larger, rather than Chinese models being 10x more efficient. It is said that GPT-4 is already a 1.6T model, and Llama 4 behemoth is also much bigger than Chinese open-weight models. Chinese tech companies are short of frontier GPUs, but they did a lot of innovations on inference efficiency (Deepseek CEO Liang himself shows up in the author list of the related published papers).

GPT-4 was likely much larger than any of the SOTA models we have today, at least in terms of active parameters. Sparse models are the new standard, and the price drop that came with Opus 4.5 made it fairly obvious that Anthropic are not an exception.

man i miss GPT-4.5

Re: No, it doesn't cost Anthropic $5k per Claude Code user

#320

Post by root_axis » Tue, Mar 10, 2026, 8:36 PM UTC

elbasti wrote: post #280

"Any conversation about token costs devolves into an ad-hoc, informally-specified, bug-ridden implementation of half of generally accepted accounting principles." We have a way of determining if Anthropic is, or has the capability of being profitable, and what the levers to that may be. AI may be world-changing, but the accounting principles behind AI labs are no different than those behind a Pizza Hut. Even if the c…

> They could find a way to climb up the value chain and capture more of the consumer surplus.

Yes, this is exactly why OpenAI and Anthropic are hyping AGI. If LLMs ever become good enough to replace workers, the first sign will be frontier model companies launching competitor businesses. It doesn't make sense to sell the formula for gold when you can just use it yourself.

> There could be a paradigm shift in compute architecture/compute cost.

Possible, but no signs of this on the horizon. If it does happen, it's impossible to predict when it will.

> We could reach a limit of marginal utility, shifting consumption to legacy models, thereby lengthening the depreciation/utility of training.

I'm not sure market dynamics will allow this any time soon. We seem to have already achieved a marginal utility equilibrium in terms of model size, so training new models on trending use-cases (e.g. synthetic data targeting tool calls, agentic workflows, computer use, etc) is really the driving force behind product differentiation. Nobody wants to admit "training new models isn't profitable" because that deflates the AGI singularity narrative that all this investment hinges on.