The performance of local AI for coding, given correct setup & harness, is very good and on par for 95% of use cases compared to closed models.
Local usage for end-users is unmetered and token speeds are very fast and keep improving every few weeks due to optimizations being done in the community.
Data remains local which makes EU compliance checkboxes way easier since data never touches anything related to the CLOUD act.
While I don't know your specific use cases, when interpreted generally, this seems trivially disproven by putting eg. websites made by Opus 5.5 and the local model side by side.
What's the problem?
Frontend has been solved for years.
With GPT-5.4 or GPT-5.5, or of course also by writing the code by hand if we want to go all the way, it was well possible to build nice frontends to our liking.
I still don't want to go back when I can do it better and faster with Opus 5.5.
I am geninuely interested what stack you are using here. Could you please elaborate?
It feels like things changed in the past few months. We are now truly able to do what we need to do without Claude etc, where this was not feasible earlier this year.
We're back to early 90s prices.
EDIT: Yes, obviously you don't need a RTX5090 to play games. But it seems to me like the top GPU market is now exclusively for the local LLM users.
An RTX 5060 Ti 16GB card costs about $800 on Amazon right now. Running 18.4 hours a day ( off peak electric rate schedule ) it can process approximately 109.3 million tokens per month. The card has a 90 day return window. Credit card interest + electricity puts the cost at $80 total out of pocket, to use this card for 3 months and return it.
Paying an AI provider came out to $1,639.50 for the same 3 months of 18.4 hour a day usage.
You know its a bubble when ... $80 vs $1,639.50
I understand that if you're a major corporation, or a researcher that needs 100+ gpus for a large project, unless you have access to your own supercomputer, the economics will be different.
This was for "individual coder at home" scenarios.
tl;dr ... im exploiting jeff bezos, and math is racist. too funny.
I guess people have never heard of "options" before.
I optioned a video card for 90 days. My bad.
On the other hand, if you do include the cost of the hardware it doesn't seem change the overall conclusion.
They gave me a 90 day evaluation window.
They did this because it benefits them.
I paid 35% above retail for the card.
And I'm the greedy one ? really ?
Thank you for contributing to current year.
MBA school calls it responsible corporate stewardship.
You're counting on most people being willing to abuse a return policy like that. Also, the 3 month return window must've activated 3 days ago. These prices didn't start just then.
In fact I'm running qwen3.8-flash-next with a Playwright-mcp docker and it is far better for scraping than my ChatGPT or Claude subscription, for the simple reason that sites aren't lying to Playwright.
Not to invalidate your point, just saying it's a bit of a straw man.
The 5060 wasn't the fastest/$, it just provided the most VRAM / $ as of yesterday.
Im installing it in a box with a 4070 Ti to run 2 different models simultaneously. Coder + MoE via paperclip or multica.
That probably is the cheapest per 1 GB vram right now. However, the R9700 a few months ago was only $1350/32GB or $42/1GB VRAM. And everyone knew it, because the card is now $1800 or so with supply issues.
I think the market discovered the reasonable min is 32GB. That runs Qwen3.8 Q6 with 100k+ context all in vram. 24GB does not.
localaicost.com
What it does buy, though, is privacy and freedom (nobody can decide to "change the terms" to milk you more).
- rate at which tokens are generated
- quality of generated tokens
- ethics of abusing return policies
- long-term viability of abusing return policiesYou can also keep it and return a bag of flour weighing the same.
Or even go to people houses and rob them at gun point to steal their hardware.
The exact same ones are now pushing $300/ssd and Samsung's (which was $300 18 months ago) is now $500.
The prices are insane.
Also even if you can somehow fit a few 100 billion parameter model(like gpt 6 luna or deepseek flash) in 16GB, and magically assume you can run at 200 tokens/sec according to your calculation, the blended cost of it would be around $100, not $1639.
It’s akin to saying “the new iPhone costs $2.5k, it weighs 250 g, I can get a piece of scrap aluminium for $30, checkmate liberals.”
I too did a cost breakdown of buying my GPU for ~10K EUR more than a year ago. It came out "Invaluable" as I use it for analyzing my diary, financials and other private things I cannot send to random 3rd parties.
But even the best of models you can run with 96GB VRAM really doesn't come close to the hosted SOTA models, sadly.
What scam are you speaking of ? The return window ?
I don't plan on returning it, unless the economics of upgrading in 89 days favor doing so.
The point was to calculate "my risk" on the transaction.
As for the results of the calculation, I didn't invent math or risk.
For $80 total "risk", I can get 10+ programmers worth of work for 3 months. A much better gamble than paying per token.
This seems more like a sociopathic "hack" than an actual comment on the economy of selfhost vs inference-as-a-service.