GitHub Copilot Users React To New Usage-Based Pricing System (arstechnica.com) 74
An anonymous reader quotes a report from Ars Technica: In April, GitHub announced that it was moving subscribers from request-based billing to a usage-based model for its AI-powered Copilot service. As that new pricing model goes into effect today, many GitHub Copilot users are reporting some extreme sticker shock as they realize just how quickly their previous "normal" usage is burning through their newly limited monthly allotment of AI credits. Across social media and forums, many Copilot users are sharing personal statistics showing how just a few hours of AI usage can now account for a large chunk of their new monthly subscription caps. For some users, it reportedly took less than a day to use up a month's usage quota.
That's a big change from previous months, when GitHub Copilot subscribers were allocated a certain number of "requests" and "premium requests" based on their payment tier. GitHub said that the old system meant that "a quick chat question and a multi-hour autonomous coding session [could] cost the user the same amount," forcing Copilot itself to "absorb much of the escalating inference cost behind that usage." [...] Indeed, some Copilot users have been sharing estimates from GitHub's own tool showing that their previous monthly usage would rack up bills in the thousands of dollars under the new pricing plan. Under GitHub's new usage-based pricing system, paid Copilot subscriptions instead grant users a certain number of AI "credits" each month, with one credit corresponding to $0.01 of usage. Subscribers also get bonus credits depending on their subscription level: the $10/month Pro plan includes 1,500 credits ($15 worth); the $39 Pro+ plan includes 7,000 credits ($70 worth); and the $100/month Copilot Max plan includes 20,000 credits ($200 worth).
The precise number of Copilot credits used by a given prompt is determined by the number of input and output tokens used and the rates charged by the underlying large language model. That means pricing is highly dependent not just on the type of request but on the specific model that a user chooses. One million output tokens from OpenAI's GPT-5.4 nano would run just $1.25 on GitHub Copilot, but that same level of output would run $30 on the frontier GPT-5.5 model (Copilot users who rely on "Auto" mode to pick the most appropriate available model for any request should be extremely careful, as some users report it can switch to expensive models for extremely simple queries).
That's a big change from previous months, when GitHub Copilot subscribers were allocated a certain number of "requests" and "premium requests" based on their payment tier. GitHub said that the old system meant that "a quick chat question and a multi-hour autonomous coding session [could] cost the user the same amount," forcing Copilot itself to "absorb much of the escalating inference cost behind that usage." [...] Indeed, some Copilot users have been sharing estimates from GitHub's own tool showing that their previous monthly usage would rack up bills in the thousands of dollars under the new pricing plan. Under GitHub's new usage-based pricing system, paid Copilot subscriptions instead grant users a certain number of AI "credits" each month, with one credit corresponding to $0.01 of usage. Subscribers also get bonus credits depending on their subscription level: the $10/month Pro plan includes 1,500 credits ($15 worth); the $39 Pro+ plan includes 7,000 credits ($70 worth); and the $100/month Copilot Max plan includes 20,000 credits ($200 worth).
The precise number of Copilot credits used by a given prompt is determined by the number of input and output tokens used and the rates charged by the underlying large language model. That means pricing is highly dependent not just on the type of request but on the specific model that a user chooses. One million output tokens from OpenAI's GPT-5.4 nano would run just $1.25 on GitHub Copilot, but that same level of output would run $30 on the frontier GPT-5.5 model (Copilot users who rely on "Auto" mode to pick the most appropriate available model for any request should be extremely careful, as some users report it can switch to expensive models for extremely simple queries).
But you MUST love the gen AI bots! Or... (Score:2)
Why the FP brain fart? Or am I just reflecting too much.
I actually started using GitHub for a new project, but without the Copilot seasoning... Which somehow led me to these rants of the day?
So how do y'all feel about Microsoft's new Copilot websearches? It seems to me that the successes are minor and forgettable while many of the failures are spectacular far beyond merely being wrong. The better to sell more advertising? I wonder if some AI can explain to me how this makes economic sense as the stock marke
Should get really exciting. (Score:5, Interesting)
So long as they were losing money hand over fist the vendor does want to throw enough tokens at you to make you feel like you are having a good time; but as few as are required to do that since they lose money on every one. If they were breaking even or turning a profit the incentive would be to sneak as much spend and upsell in as possible; and it's well known that the verbosity/cost of LLM chatter is hard to predict; harder if there are multiple models and other complications being switched around in the background.
What sort of exciting little tricks will we see from vendors who actually make more if you use more?
Re: (Score:2)
new material? the reason ai porn is so prevalent is because it is pretty much the same for 100's of years if not longer.
Re: Should get really exciting. (Score:2)
Now you can mix and match parts.
Re: (Score:2)
So, this is in the process of being regulated as gambling?
Re: (Score:2)
Re: (Score:2)
It's the equivalent of Netflix saying they are going to per video watch pricing.
It's worse than that; it's like they're basing the price on how long the video is and what kind it is - a 90m romance vs a 180m sci-fi. Perhaps giving you a credit if you can sit through an entire Michael Bay flick w/o wanting to kill yourself. :-)
Re: (Score:2)
Not sure what you mean by the last sentence. Acceptable to whom? It makes it sound like it's a great matter of principle. But I don't think there is one here.
An overly hyped technology, promoted using lies and exaggerations about its capabilities, built through fraud and IP theft, designed to be addictive by carefully crafting its output to stimulate dopamine production, marketed to CEOs who use it in place of cocaine of yore and believe the hype because of an inability to listen to those warning them, impo
Re: (Score:2)
Good news: every subscriber was informed of the changes, and had opportunity to cancel / not use it if they didn't like the changes. In fact, I even hear there are multiple providers of AI services for developers if they want to change providers to take advantage of cost competition.
This should surprise nobody, because they already know how many tokens they're using if they are paying attention at all. They could have done the math to figure out what last month would cost under the new structure if they w
Re: Should get really exciting. (Score:2)
First rule of capitalism, if people are paying it then it was acceptable.
Re: (Score:2)
Should that hold true for more people I wouldn't bet much on AI adoption.
Re: (Score:2)
More than that. The monetization strategy was their only key differentiator and now there's no reason to use it beyond making the nag screens to stop.
Re: (Score:2)
I think they will need to raise prices a lot more to get to "losing money".
True cost of AI LLMs (Score:4, Interesting)
Hopefully AI thing will start to cool a bit now that the free ride is over and end users start to bear the full cost of these systems. Up until now the big AI companies have been giving it away at a token cost to keep the hype going.
Hopefully seeing the true costs will push demand more towards more efficient, smaller, specialized local models for many applications.
Re:True cost of AI LLMs (Score:5, Insightful)
"Cool a bit"? If the general truth about the subsidizing of prices gets out, we are looking at a big bubble burst at least as bad as the dot-com poppage.
Investor funds and market-share-fights have kept AI prices low or free, but of course that can't last forever. I suspect one prominent but stressed AI company will spill the beans about fake pricing ("we all do it!"), putting pressure on the rest to prove that claim is false, which they'll fail, spooking investors, ending the run, and triggering a recession.
Re: True cost of AI LLMs (Score:2)
Bailout incoming
Re: (Score:2)
Massively worse. The dot com boom didn't hurt existing businesses except maybe the occasional telco that saw a drop in subscribers to T1 lines.
Major businesses have restructured themselves around AI and essentially started losing control of their business knowledge. They have CEOs that have gone all in on AI thanks to slick marketing and dopamine-based AI design, and will now have to admit they made terrible errors of judgement, which is something most aren't even going to admit to themselves, in doing so.
Re: (Score:2)
Gambling on stock futures requires more than an understanding of what will happen, it requires knowledge of WHEN it will happen.
I agree with the OP on eventual outcome, but that doesn't mean shorting a specific stock now is justified. Nice troll though, a classic technique for those who pretend to be knowledgeable.
Re: (Score:2)
Yes. That is by far the most likely outcome and it may not happen basically at any time.
Re: (Score:2)
AI isn't merely about selling investors on a profitable service, AI is selling investors on the creation of a "greater than human" intelligence that will transform life as we know it. Investors need to own a piece of the enslavement and trivialization of humanity.
Spooking investors may well happen and this prediction may prove to be true, but it may very well not happen that way and we may learn that this prediction turned out to be best case optimism. If only a bubble popped and a recession triggered.
Re: (Score:2)
These price increases are nowhere near the "full cost of AI". For that they need to go up like 10x or more of the prices before this price hike. Remember that they currently burn about 80% of the money invested and that they have been doing that for a few years. And they need to make up for that.
Turn off automated PR reviews (Score:2)
I'm disabling all automatic PR reviews for sure. I was going to do this anyway, but it was a setting the org controlled so I didn't have a choice. Now I think we have an argument for ending Copilot's review feedback. Because while the models are fine at reviewing code, mostly, the Copilot tool itself is garbage.
Post Customer Acquisition (Score:4, Insightful)
Now who is surprised that prices go up after the customer acquisition phase?
Memory and compute are expensive. Data centers are aren't as easy to build as once thought. [finance-commerce.com]. Power is expensive. People fights over land and water rights.
And if the rest of the pack like Claude [theregister.com] moving to usage based limits. Then might as well make the change, Microsoft isn't a startup which is burning investor cash. They are definitely going to make a profit. AI ain't cheap and they already have the customers.
Re: (Score:3)
Now who is surprised that prices go up after the customer acquisition phase?
Like all drug dealers, the first hit is always free.
They should have waited until more jobs were eliminated and companies were more dependent on them.
I switched to Cursor AI after Copilot rolled out the 300 request cap. It is vastly superior overall, and whatever type of account my company provides, I've yet to hit a usage limit.
Re: (Score:2)
Re: (Score:2)
Yes. And they will have to increase prices by a lot more to become profitable. All of them will have to. There are already companies that find even moderate price increases will make human devs cheaper than AI, though.
Re: (Score:2)
So long and thanks for all the circular funding (Score:3)
Let's recap. First, VC and mega-tech corporate coffers were the sources of unlimited funding. Compute was bought and essentially offered free or nearly free to even the most voracious users to juice usage stats and hopefully disrupt that too-long-entrenched habit people have of, let me check my notes, "using their brain to think." And to do 30% of the job of a person at 5% of the cost, sure, why not! Four to six months ago, those funds started drying up.
Fortunately, there was circular AI "investments" and "deals" between Nvidia, OpenAI, Microsoft, etc, which of course wasn't funding at all, just moving money between multinationals/hyperscalers to double- or triple-count the value of the money as it changed hands and maintain the air of inevitability by sheer scale of AI investment. Sure, it wasn't real value, but it was good enough to (maybe, if SpaceX, OpenAI and Anthropic move quickly) give runway for a few IPOs, and heck, when we're only burning through a few billion dollars a month, what's the harm in placing a bet to own the all future labor value?
But now? It looks the CEOs, boards, and shareholders remember they like their present-day selves more than their future-selves (future me? That guy is basically a stranger!), who are really the beneficiaries of these costly bets. So, with no other funding sources, AI companies have no choice but to charge users for compute directly. So now AI can do 30% of the job of a person at 10,000% of the cost.
I think we all see where this is going.
Re: (Score:2)
that too-long-entrenched habit people have of, let me check my notes, "using their brain to think."
[citation needed]
Re: (Score:2)
Alternatives (Score:4, Insightful)
If AI services are becoming too expensive in the current environment, we can look to nature for help. There is a an abundant species of large mammals in the ape family that can be trained to do this kind of work as well.
Re: (Score:3)
Reminds me of The Feeling of Power [wikipedia.org] by Isaac Asimov.
Re: (Score:2)
At this point a man called Gardrilla Manceframe rediscovered and patented a device he had seen in a history book called a staircase. It has been calculated that his most recent tax bill paid for the social security of five thousand redundant Sirius Cybernetics Workers, the hospitalisation of a hundred Sirius Cybernetics Executives, and the psychiatric treatment of over seventeen-and-a-half-thousand neurotic lifts
When confronted with the Sirius Cybernetics elevators ("Happy Vertical People Transporters").
Re: (Score:1)
But they smell like stale Cheese-Whiz
Re: (Score:2)
Dune's Mentats.
Re: (Score:2)
Even at the more accurate prices it's still cheaper than paying a six figure salary. So while we will probably see a lot less AI doing peoples' hobby projects, or even open source projects, companies are not going to reduce their internal use of it.
Re: (Score:2)
That's what the AI companies count on, wonder whose stock you're invested in?
Companies don't have to "reduce their internal use" if they don't have any. It remains to be seen of there will ever be a critical "mass" of corporate usage to make your point interesting. Meanwhile, models are available, the cost is only cloud computation. If costs to corporate customers is high, they can run models locally.
Re: (Score:2)
The tasks AI is capable of are not six figure salary tasks.
Extract data from an emailed AP invoice attachment and let some low paid grunt review the metadata for approval and applying corrections. Saves manual entry time. Maybe it can cut the cost of processing an invoice from $5 to $2 per invoice and at scale can help free up the AP Admin or keep from hiring additional staff. Currently AI is decent enough for assisted back office tasks. If the prices of the AI tasks increases like we are seeing then the sa
Re: (Score:1)
If AI services are becoming too expensive in the current environment, we can look to nature for help. There is a an abundant species of large mammals in the ape family that can be trained to do this kind of work as well.
You'll never legally hire an employee for $200/month.
Anywhere with access to Internet services has laws against that.
This is the core reason we accept the poor quality results from AI far easier than from a human.
$200/mo for a program vs 2x paychecks, benefits, taxes, DOL escrows.. Plus all the other accommodations for humans most employees don't consider.
One should expect far less in results when paying far less.
Of course one should also always do a cost/benefit and apply some reason to what can and can't
the point is (Score:2, Interesting)
It costs far far more than $200 a month to use AI in the way people are doing. The people referred to in the article (I'm assuming you did not read it, in the fine tradition of Slashdot) are at something like $13k of tokens used per month. And since the AI is only an adjunct to a human brain, even if it doubled the throughput of an engineer it would not be worth it, and of course it does not double productivity.
AI is absolutely amazing for some things, even finding bugs ("the app crashed when I did X, I thi
Re: (Score:3)
Re: (Score:2)
1. Company uses AI applicant tracking system to weed through resumes and applications
2. Applicants use AI to create resumes and applications designed to defeat ATS filtering
3. AI companies make money coming and going,
4. Companies / applicants enter a vicious circle of cat-and-mouse AI dumbfuckery that ultimately turns electricity into heat as the only recognizable output.
Progress!
Well, that's a shock. Not. (Score:4, Informative)
It's the classic enshittification cycle, also beloved by drug-dealers.
1. Get users hooked on free or cheap service.
2. Once they're hooked, jack up the price.
Re: (Score:2)
In the AI coding space there are many alternatives. And, as reported, this is a huge change- from $10/month to $10 a day. They're not boiling the frog - creeping prices up slowly so no one quite notices.. Watch for the unsubscribes - mine included...
Re: (Score:2)
I take it you have never actually met a drug dealer. Perhaps you should try allegories that are based on reality.
Re: (Score:2)
Exactly. And as they are all currently burning money like crazy, all the others will increase prices too.
Re: (Score:2)
Not everything you don't like is enshitification. It's not even obvious that this is making anything worse, just making things realistic.
Know your taxonomies!
Someone has to pay for it (Score:3)
Comments on here repeatedly say people should only pay for what they use, whether this be water or electricity. Using AI is no different. The more you use the more you pay.
Not sure why this is controversial or surprising.
Re: (Score:2)
Because most libertarians interested in pay-go schemes are interested in OTHER people paying. Same as it ever was.
Re: (Score:2)
Not sure why this is controversial or surprising.
The surprising thing here is what the actual cost is. People have been completely and totally insulated from it until now as every AI company was effectively acting as a loss leader and no one in the industry was charging sustainable rates.
The system is starting to financially collapse, and maybe now that we start seeing the true cost we can back down from this insane bullshit path we're on.
Re: (Score:2)
The surprising thing here is what the actual cost is.
Why is that a surprise? We all predicted it from the very beginning. It's like being surprised when Christmas arrives every year.
Re: (Score:2)
Not sure why this is controversial or surprising.
Probably because these people assumed in their boundless naivete that AI will remain cheap. My personal estimation is that they will have to increase prices by at least 10x to start to become profitable. I may be grossly underestimating this though.
Re: (Score:2)
"Not sure why this is controversial or surprising."
Because it is neither and no one says otherwise. Great takeaway from you though, are you in middle school?
How much per token? (Score:4)
The problem with all these pricing models is that the concept of a "token" and what it consists of is nebulous at best. I know what they mean by it, but there's not hard calculation to determine how many "tokens" a specific prompt or response will consume. You can guess, but only so accurately. That concept needs to be better defined before anyone can estimate their ongoing costs.
Tokens are stupid, gigawatts are where it's at. (Score:3)
Re: (Score:2)
AI companies don't want customers estimating ongoing costs.
Re: (Score:2)
A token is actually well defined. The problem is that it isn't exposed.
It's time for Github'ers to Move On (Score:2)
Microsoft's Strategy is Backwards (Score:2)
Re: (Score:2)
The IT department where I work just enabled Copilot in Office with great fanfare like it's going to save us so much time.
I literally only open not-Teams and not-Outlook maybe once a week, to read a document someone else sent me. The integration in Outlook appears to just give me pop-over bullshit about how it can try to find me more tasks to do in my shit show of an inbox I ignore as everything important comes through Teams.
Quite frankly, I don't know what we're paying for, other than automatic transcripti
All of the AI companies have mountains of debt! (Score:2)
It remains to be seen if the users can afford the Automation as a Service that these companies are trying to sell.
Everything these AI companies have at this point, is heavy on the Artificial part and completely lacking even a hint of the Intelligence part.
about time (Score:2)
welcome to the actual cost of AI compute, people. this price is still probably subsidized, wait until they start charging a profit-generating price.
Iâ(TM)m glad to see it, the climate impact of free AI powered by fossil fuels is devastating. if everyone started charging the actual price for the compute required for AI, that would help move us out of the fast lane to climate apocalypse that weâ(TM)re currently in.
I'm switching to using GPT-5.5 exclusively (Score:2)
Maybe the cost will finally persuade my employer will stop shoving CoPilot down our throats.
Cancelled my Copilot subscription yesterday. (Score:2)
The price correction wave has begun.
Guess what, suckers... (Score:2)
...getting hooked on the stuff is free. But then the bill comes due. And the prices will have to go up a lot more than this small increase to make the LLM peddlers profitable. This price hike is just what they think they can get away with at the moment.
Dependence (Score:2)
By the time they jack up the price the brainrot is already too far gone. You don't remember how to code anymore. You have no choice but to keep paying for tokens. Eventually you end up begging on the streets so you can afford more tokens. You rob a liquor store so you can pay for tokens and end up going to jail, where you go through rehab and are finally free of the need for tokens. But when you get out, you can never quite get rid of the desire for more tokens, always in the back of your mind... This is ho
I can help define this (Score:2)