Interesting, I didn't know minutes where free before.
Stopped my recurring subscription at the end of last year when it started spinning up actions for review. Which as a side effect doubled the time (or so) to do a review. Whereas before that I would open a PR, wait at most a minute or two and the review was already done.
Expect to see more of these kinds of announcements as companies need to start showing returns on their AI investments. It's hard to say how subsidized the current AI products are[1] but we're definitely getting a free lunch at VC's expense the moment.
[1] Ed Zitron speculates the actual prices with token based billing for heavy users will be something like 10x the subscription price, but this seems high.
Good thing GitHub has plenty of built-up goodwill to spend down. If they didn’t this cascade of (probably necessary but nevertheless negative for customers) changes might be the tipping point to push a lot of companies to seek other options.
GitHub has already lost so much good will. It was one of the best pace on the internet for like a decade and is now just a real pain in the ass to interact with
Unclear why this is so shocking. Sounds like they have been making migrations on their underlying systems and this better aligns with the cost to run. I would be curious how many are using their code review system.
is there any data on how many Actions minutes a single copilot review actually takes? the announcement doesn't mention it, and for a team doing 20+ PRs a day that number adds up fast.
They should also remove Copilot code reviews from being counted as metrics in a PR.
I've seen some projects that use it and you open the PR page to be greeted by every PR having 3-20 comments but when you goto the actual PR, there's no one except the contributor with a bunch of Copilot feedback.
It gives a false message that the PR is resonating with folks and has real activity. I wonder if GitHub did this on purpose to make engagement seem higher than it really is.
I started moving my repos off GitHub weeks ago, but I'm still waiting for a "good" GitHub competitor to appear. GitLab sucks (if you're not a company and like unnecessary complexity), Codeberg is slow and limited (and has weird mods), Sourcehut UX is weird (and being DDoS'd), Gitea Cloud didn't even have a working login page last time I tried, BitBucket isn't the worst but it has quite a few problems (& isn't set up for public repos/search). Please can somebody start a simple, reliable hosted GitHub alternative? I'd pay for it...
This is funny, but Copilot is still an interesting case-study and (probably) failed predictor of where we are headed.
We all know, and have known for a long time, that the AI labs selling dollars for a nickel are going to pull that rug, and up that price, at some point.
Copilot, though, has been consistently the weakest mainstream AI coding offering. Inferior to Cursor or Windsurf at editor completions, inferior to Codex, Claude, OpenCode, blah blah blah, at agentic coding and also the old-school chat-style...
And now, it's no longer cheap AND now sucks even more than it has all along — the new $39/month plan is not only worse than all its competitors, but worse than its own $10 plan was a month ago — by a lot.
The thing is, you can't jack the price up unless you're good enough — at least on some axis, to some customer segment — to jack the price. And when you're not good enough, and you have vastly superior competitors who are not doing that yet... you're just forfeiting the game.
Which I agree, Copilot should do — it's the Windows Phone of AI coding assistants, after all — it still seems weird to me to just commit humiliating suicide rather that trying to make some deal with one of those superior competitors.
Instead of just jumping into a dumpster and lighting yourself on fire.
I suspect Microsoft will reneg if enough people cancel.
Even before yesterday, I assumed they made money via the gym model. I'd have months where I'm too busy to use my co pilot subscription in any meaningful way.
Canceling and restarting is too much of a hassle.
But with the pricing update I'd probably use up the 10$ plan within 3 days.
I don't know if anything else is integrated so we'll into GitHub though. I might keep the 10$ plan just for the occasional GitHub AI PR.
If their pricing turns out to be what they claim, and copilot cli has accurate token counts, they had the best deal around.
Just today, when I wasn't being especially chatty with GHCP, I used about 12 requests to get a few thousand line changes in 3 projects I'm juggling. The last project repo of copilot I closed, in 3 hours burned 38M input tokens, 28M cache, and like 400K out. For GPT5.4, high. That's like $135, in half the day, 1 of 3 instances. No crazy tool use, just lots of docs and unorganized code. GHCP charged like 70 cents for that on the old plan.
Github was already struggling with bazillions of throw-as-much-crap-on-the-wall software running in actions, and now the world is running throw-as-much-LLM-crap-on-the-wall computation, as unstoppable as the pre-LLM era. Turning compute into excrement as fast as the planet is filled with it. Excrement being "Github Copilot code review" in compute world, and no need to draw what it is in our real world.
Weird that Anthropic decided to build a Claude Code Routines toilet.
We are slowly inching closer to the point where AI and AI products will be billed for what they cost. We are currently living in the heavily discounted world where everything subsidized to the point where a lot of it is free. It seems like they can't or won't keep that up anymore. My prediction is that whenever one of the big companies raise their prices or move features to higher tiers others will follow soon. They all feel the pressure and non of them want to give away more money than they need to.
I wonder if managers will be as excited about AI when the prices go up.
But with new hardware comping out, and maybe models being smart enough to help with optimizing them and reducing inference costs even more, I think we should still expect the costs to go down.
I see it as no different from the previous generation of consumer startups burning money - as Derek Thompson wrote,
> ...if you woke up on a Casper mattress, worked out with a Peloton, Ubered to a WeWork, ordered on DoorDash for lunch, took a Lyft home, and ordered dinner through Postmates only to realize your partner had already started on a Blue Apron meal, your household had, in one day, interacted with eight unprofitable companies that collectively lost about $15 billion in one year.
Wonder if folks will flock to competitors that still over fixed-price review packages only for those companies to switch to usage-based pricing. They're next to be overrun with new customers and sit on the increasing bills.
I use AI when its free, but very sparingly when I try to get some sample code. Paying for that thing seems like an insane concept to me. I'm really glad it never became integral to my skillset or day to day work. Seems like a lot of people will feel the pain of getting rug pulled harder sooner or later. At this point any company sensibly investing in this technology is going the offline/self host route, because they smell the pattern of vendor/cloud lock in ten miles away
I love LLM tech and want to see it support viable long term businesses that don't rely on losing money providing subsidized subscriptions that lose money. I ranted about this on X yesterday, primarily about a legal framework to prevent LLM dumping: https://x.com/mark_l_watson/status/2049221176657129794
34 comments
[ 2.0 ms ] story [ 54.2 ms ] threadStopped my recurring subscription at the end of last year when it started spinning up actions for review. Which as a side effect doubled the time (or so) to do a review. Whereas before that I would open a PR, wait at most a minute or two and the review was already done.
[1] Ed Zitron speculates the actual prices with token based billing for heavy users will be something like 10x the subscription price, but this seems high.
I’m blowing through my 1000 mins in days.
Thinking to either pool some free tiers or figure something out with spot instances.
Also is it just me or is CI/CD tooling still sort of rough all around.
It is time to setup local models. It is cheaper, and you already have a computer. Why keep it idle and pay someone else for their CPU?
I've seen some projects that use it and you open the PR page to be greeted by every PR having 3-20 comments but when you goto the actual PR, there's no one except the contributor with a bunch of Copilot feedback.
It gives a false message that the PR is resonating with folks and has real activity. I wonder if GitHub did this on purpose to make engagement seem higher than it really is.
We all know, and have known for a long time, that the AI labs selling dollars for a nickel are going to pull that rug, and up that price, at some point.
Copilot, though, has been consistently the weakest mainstream AI coding offering. Inferior to Cursor or Windsurf at editor completions, inferior to Codex, Claude, OpenCode, blah blah blah, at agentic coding and also the old-school chat-style...
And now, it's no longer cheap AND now sucks even more than it has all along — the new $39/month plan is not only worse than all its competitors, but worse than its own $10 plan was a month ago — by a lot.
The thing is, you can't jack the price up unless you're good enough — at least on some axis, to some customer segment — to jack the price. And when you're not good enough, and you have vastly superior competitors who are not doing that yet... you're just forfeiting the game.
Which I agree, Copilot should do — it's the Windows Phone of AI coding assistants, after all — it still seems weird to me to just commit humiliating suicide rather that trying to make some deal with one of those superior competitors.
Instead of just jumping into a dumpster and lighting yourself on fire.
Even before yesterday, I assumed they made money via the gym model. I'd have months where I'm too busy to use my co pilot subscription in any meaningful way.
Canceling and restarting is too much of a hassle.
But with the pricing update I'd probably use up the 10$ plan within 3 days.
I don't know if anything else is integrated so we'll into GitHub though. I might keep the 10$ plan just for the occasional GitHub AI PR.
It seems everything Microsoft does is like this nowadays. They just can't seem to win anymore.
Just today, when I wasn't being especially chatty with GHCP, I used about 12 requests to get a few thousand line changes in 3 projects I'm juggling. The last project repo of copilot I closed, in 3 hours burned 38M input tokens, 28M cache, and like 400K out. For GPT5.4, high. That's like $135, in half the day, 1 of 3 instances. No crazy tool use, just lots of docs and unorganized code. GHCP charged like 70 cents for that on the old plan.
Weird that Anthropic decided to build a Claude Code Routines toilet.
I wonder if managers will be as excited about AI when the prices go up.
I see it as no different from the previous generation of consumer startups burning money - as Derek Thompson wrote,
> ...if you woke up on a Casper mattress, worked out with a Peloton, Ubered to a WeWork, ordered on DoorDash for lunch, took a Lyft home, and ordered dinner through Postmates only to realize your partner had already started on a Blue Apron meal, your household had, in one day, interacted with eight unprofitable companies that collectively lost about $15 billion in one year.
Between 27x model costs and this, CVE exploits and downtime their platform is starting to feel like a questionable decision.