Reminds of the “girl math” meme that’s trending amongst millennials and younger: https://knowyourmeme.com/memes/boy-math-girl-math

I had Codex queue any CAPTCHAs for me to solve. Also set up one sub-agent to search Reddit for other people's clever approaches to getting good rental car deals. Ultimately, it found a third party reseller for Avis that had a price of $600 using a discount code for both cars vs $1,100 cheapest from Avis directly. I would not have discovered that on my own, at least not without like a week's worth of manual searching. So right there that was a $1,000 saving for both cars in exchange for 20 minutes of writing a decent Codex prompt.

Interesting, you don't find that the model is perfectly capable of solving those on its own?

Or does it just refuse to try?

I have a Grok Bot that just looks for glitched pricing and crazy deals on a list of items I need and it reports in a couple of times a day.

How else can we get these things to make us/save us money?

My tests in the past month, basically $0 saved over my findings: - Two one-way flights - New OLED 55"- 65" TV - Mini PC for CI/CD runner, with very specific specs - Expensive, specialized medical device - Variety of car parts

I think one problem is that I simply already know the places to check for this stuff. The more esoteric places (ebay/fb marketplace/etc) agents seem to just be awful at parsing what is real, what is a scam, what is a good deal, etc, and that is where the big deals are for stuff.

That would limit the $500 plan to smaller businesses.

I think this is the first instance where it means you are actually paying 500$

Worth it? Huge YMMV. I'd expect a $1000 sub next.

  Cosmo Kramer: It's a write-off for them.
  Jerry:        How is it a write-off?
  Cosmo Kramer: They just write it off.
  Jerry:        Write it off what?
  Cosmo Kramer: Jerry, all these big companies, they write off everything.
  Jerry:        You don't even know what a write-off is. 
  Cosmo Kramer: Do you?
  Jerry:        No, I don't.
  Cosmo Kramer: But they do, and they're the ones writing it off.

Or maybe it gives you just a little bit more tokens than older $200/month plan for plausible deniability:

    Maybe me: “you’re charging $500/month for what was $200/month earlier”

    Maybe OAI: “no we’re not!! You get *more* tokens than the older $200/month plan and did you hear, you get Astra Ultrafast access!”

    Maybe me: “oh really? How many more tokens??”

    Maybe OAI: “I’m sorry we can no longer share that information, it’s confidential. National security and all…”

    - $500 plan comes with not 20x,
      but 25x allowance!
    - also Astro Ultraspeed
    - $200 pulled down to 10x quota

    - one time ~62,000 credit
      top-up ($2,500 API call 
      value)
    - 5 hour quota limits removed

Maybe OAI thinks people are creating new unnecessary work to use their allowance completely. Well they aren’t entirely wrong. But the experimentation has been useful in letting us discover new use-cases, and to develop and flex our agentic muscles.

Dev salaries are high and devs often have to realign buttons or change colors. We are not always working on complex ideas.

I rewrote a big piece of software from scratch....it had 0 docs. Documenting of classes and flow took me weeks. Complete rewrite took almost 2 years. I am confident that with AI I would be able to get this done in a few months.

Gets to put you under pressure to deliver more while paying you less. A few months you say, why not weeks? I'm sure the latest model can do it.

Also, forget about those "easy" ideas, if you are not on the verge of burnout every day you might as well not be working.

And if they do that accurately, there's no stopping someone from mapping the guard rails.

Clearly that outcome is more uncomfortable than what we have now, which is enough people seem to be okay with vague billing.

Second, how does vague billing prevent someone from mapping the "I'm sorry, I can't help you with that" style of responses to map the guardrails?

The API makes it simple, yes. They make you pay per token per model. A new model has different prices. Not being able to do that is what makes it not so very simple.

Just trust me bro accounting.

Maybe one day open weights and closed models will hit the same wall?

Overall Ai is doing all the software development now and we the consumers are paying for this while many are out of work and if you are out of work the US tech job market is brutal!

Ai needs to be free to use without limits and in time Im guessing it will be. That could be Meta and or another who accomplishes this. Whoever does wins this race.

Previous $200 tier: 20x pro value

New $200 tier: 10x

New $500 tier: 25x, plus ultrafast mode but it drains 6x the cost (!!)

The $200 cut to 10x was announced only formally on Twitter

However none of their product pages mention concrete #s, so I guess value will only degrade more over time.

It's strange if that ends up to be this way. It's hard to see why would anybody subscribe to the $200 - or even the $500 for the sole benefit of using ultrafast (depleted probably within hours).

Or it's saying it's always on fast mode therefore it's like running a 20x on fast all the time?

Super confused.

500$ is a old 200$ plan plus an option to drain it absurdly fast?

What is the value here of a 500$ plan I can't figure it out.

Yes.

> What is the value here of a 500$ plan I can't figure it out.

If you have unlimited money, it lets you give it to OpenAI to get faster results, presumably because you don't have unlimited time.

But past few weeks, I don’t think it was accurate since my usage was going down faster. They nerfed the usage and now aren’t even giving estimates.

That aside, all of the ambiguity around usage is killing their credibility (among other things). Usage limits feel like they swing wildly day to day and week to week, and it's hard to trust any of the other claims they make when that's the case.

"Yesterday you gave me 10 Frooby Bucks and charged 2 Frooby Bucks per banana. Now you gave me 5 Frooby Bucks and charge me 1 per banana!"

Half the usage for the same price.

This is the same nonsense Anthropic got burned for. I think I'm gonna bite the bullet and find a GLM or Kimi provider already.

actual lol

Enjoy the Lyft v Uber era while it lasts!

It's getting shitty, I agree, but still feels like a net improvement to where we were over a decade ago.

But I'm still for investing in local models.

I'll cancel my OpenAI plan after this billing cycle. Let's see how this plays out.

Does anyone see the usage limits? I don't see anything indicating how much usage is in each tier.

I checked:

* https://chatgpt.com/pricing/

* https://help.openai.com/en/articles/9793128-about-chatgpt-pr...

* https://chatgpt.com/#pricing (signed in plan upgrade page)

I tried searching for "Synthetic AI Subscription" and "Synthetic Data" is such a massive keyword in AI pre-training circles it's impossible to find the company.

/rant

What is the companys URL?

If they're simultaneously cheaper than token rates but couldn't they have sold these tokens in the first place on a platform like openrouter?

And the platforms that are hosting these tokens on openrouter aren't really earning any money at all in the first place as that itself is a race to bottom (and its more so just neutral, neither positive nor loss or barely profitable) so unless synthetic is actively making loss, and if that's the case, then it might not be sustainable, I don't quite understand the economics behind it.

Is the economics behind this the fact that since its a single subscription instead of token based pricing. Some people might use less tokens than what their subscription costs and so it offsets/subsidizes the other set of people who need more tokens.

But then, this means that as you had suggested in the first place, it would be a good deal for the people who want more tokens, so they would flock even more to the platform and it should eventually balance out unless external money is being burnt as I had said earlier.

I am genuinely interested in how they make this work so I would love if someone can shine some more light to it as I would like to read more about its economics that makes it work, thanks in advance! :-D

- The way they pace and limit usage seems to echo some strategies I've seen to reduce the impact of burst usage

- They cut down the maximum context length of most models

- They claim to have some custom sauce that improves how they handle API requests and reduces error rates

- Their explanations appear focused around optimizing concurrency in the same way that the 'flex' mode of some APIs does, though I've not seen specific docs of them claiming that this is part of how they save on costs

There may be more that I haven't seen or noticed. I don't work for or with them.

So a 20x plan, that was formerly $200, would have been $4000 in API credits. Which sounds like a lot but the API is absolutely horrendously expensive - like you'll burn through $100 in 45 minutes of 5.6 Sol on High reasoning. I know this because I did this once.

I don't see any of those terms. Not even on the signed in page to upgrade to to a different Pro plan.

The plans I see are labeled, "Pro 100", "Pro 200", and "Pro 500" with the number corresponding to the price in USD.

Starting October 30, 2026, your included usage in ChatGPT Work and Codex will decrease from 20 times to 10 times the ChatGPT Plus allowance. Your GPT-6 Pro limit in Chat will decrease from 200 to 100 messages per week. Your price remains unchanged and you’ll keep your current limits through October 29, 2026.

And:

For those who want even more usage and speed, the new Pro 500 tier offers our highest usage allowance, with 25 times the usage of Plus. It also includes access to Ultrafast, with up to 8 times faster token generation.

https://this.os.isfine.org/blog/posts/chinese-flash-models-a...

100 = € 103,-

200 = € 229,-

500 = € 509.99,-

Besides the naming being insane, the pricing strategy for each seems to have been made by a different team.

So expect further downgrades in the plan features for the same price.

It looks like they have literally chosen the most straightforward sounding plan names they could think of.

This is just the beginning. Migrate to open models ASAP.

Apparently they don't want to compete now that they're doubling the costs.

Makes perfect sense.

(If you're an idiot)

I know people who have been at this as stage since around 4o with their own custom systems

OpenAI announced they built clones of apps that already exist, codex for normal people, and 2.5x price increase for the product that is already getting decimated by competition.

What is happening at OpenAI?

I can't imagine OpenAI not backtracking this unless Anthropic does the price hike too.

Same here, but I’m worried it’s the latter. We’ll probably see an Anthropic price hike in a couple of weeks.

If ANT does the price hike, OAI will be forced to backtrack because they're just not equal, OAI is behind. ANT would still be a much better value overall, and people will remember that it was OAI that made their ANT subscription more expensive. So the hiatus will happen. There's a lot of users that have $200 plans on both ANT and OAI. $500 leaves the slot for just 1 sub for them.

But I think the price hike is not going to happen. $500 is much, much harder to justify than $200 for the same usage. This is a horrible business decision, so Anthropic may just watch OAI bleed customers.

Had both a anthropic and openai then switched to 2 openai for astra. They want to capture the straddlers between both.

But this is a horrible decision. They could have eased customers into this instead of massive change. The are literally more than doubling the cost for no value add.

A smart company would have done this when they released astra or only allowed astra on 500$ plan. Or made it more sutble. I would likely have ate a 20% increase but this is over double.

I mean, other than Yegge. He seems like a pretty extreme outlier?

> After that date, your subscription will move to the lower included usage allowance. Your subscription price stays at $200/month.

Imagine buying this shit for a year, upfront, and then having to deal with this. Nice.

Subscription pricing will eventually compete with labor costs

Got it, calendar updated: switch back to Claude on the 29th.

I realize this may be an unpopular opinion, but I feel the FTC should start looking at things like this. If a well funded company can bait the market long enough to drown most everyone else out then double their price overnight (ok, 30 days..) doesn't competition just become a matter of who has the deepest pockets?

Tomorrow's FTC perhaps.

Edit: I meant its not worth it just for Dots Personal assistant ,also comparing to previous pricing of $200 - For $500 you got previously 5x more than previous $200 Pro pricing (which was 20x now is 10x). I can have real people who can drive or do physical deliveries etc, just on the delta.

It's crazy that people will have to pay 500 USD for only slightly more (and faster) usage of basically what I already enjoyed for the past few weeks. And even I got close to hitting weekly limits doing regular development across a bunch of codebases (admittedly it was on Astra High).

I feel like even the 500 USD subscriptions are not "get out of my face with those limits for all single developer use cases" (not even that many overnight tasks, I might have run a few here and there, and not even full 8 hours of work a day), but soon you'll have people needing multiple of those 500 USD subscriptions, unless they pull off crazy token efficiency gains for how the models work.

Guess the economics are catching up to AI companies, sucks that I don't have a bunch of money to burn every month.

Or are they simply lower the batch sizes on whatever hardware they are currently running?

I want them to succeed as a business so I can use their products, but trust does erode over time. I wish they would realize this and value their customer more.

The performance changes between models are almost static. Recently, GPT has shifted almost entirely toward token efficiency.

Astra is good at making 3D assets, but realistically its coding was not as good as GPT 5.6 Sol.

Have we reached a point where getting better at one thing means starting to get worse at another?

This announcement is the worst one yet.

Nerfing existing usage A more expensive top-tier plan Showing off how slow Dots is right in the demo video And GPT Space—are they telling us to use it instead of Notion or what...

Overall, throughout the entire announcement, everything OpenAI wants me to look forward to just leaves me, as a developer, disappointed.

I don't see why a company charging more for less is a surprise or would lead to thinking that AI models have reached their limits. Thats an incredible claim and it would need incredible evidence. Astra is much more capable than Luna. Luna being 'good enough' for most of my daily work just shows how strong Luna is, not that Astra isn't much more capable.

As models get more and more capable, they will only appear to be at the extremes. You'll need to see them solving more Millenium problems to notice a difference as the tide is continuously raised for all models as well.

Tomorrow we are re-opening the Pro $200 subscription

https://news.ycombinator.com/item?id=49889306

Fuck that shit!

Pro 100: 5x usage, $100

Pro 200: 10x usage, $200 (previously was 20x usage, $200)

Pro 500: 25x usage, $500

So, essentially scaling is equal. You get the same amount of usage from 25 ChatGPT Pro accounts as you do from one Pro 500 account.

It's fair, but a big of a rug pull considering how good of a deal Pro 200 was.

Now it makes more sense to get 100 plans from both Claude and ChatGPT plan to get diverse models.

It's like OpenAI wants me to give them less money or they are constrained on compute and need to shed load.

What they’re doing with billions of tokens every week, beats me. But these users absolutely do exist.

If it was a 50x $500 plan it would be a different story.

6.1 will probably perform even better.

Opus 5.5 is better in that regard than Fable 5.1 except for some bugs with more complexity which it was uniquely suited to.

Do things in a sustainable fashion and treat these models like the swappable commodity that they are by using them via abstraction layers. Bonus point is that you might be surprised how much more you can get for your money by intelligently optimizing specific model usage.

Back then the trips were subsidised by VC money, and also the drivers were contractors, so far lower internal costs for Uber.

“Over time, we see prices go low enough that it makes sense for most to buy usage as needed without there being a significant gap between what you get in a subscription and what you get in the API for a dollar spent.”

Claude has the same personal-professional split. Free, pro and Max have trained by default since Anthropic's policy change in August 2025. API and team plans didn't. Choosing between one and another is also choosing a training default. The subscriptions offers the opt-out setting, Data Controls for ChatGPT and Account settings for Claude, and it's worth checking before paying.

Be a better friend to yourself. Value your time more.

1. In this absolutely wild era of model progress, with frontier and near frontier models leapfrogging themselves seemingly on a weekly basis it's good to get experience with a variety of providers

2. This isn't exactly hard with OpenRouter!

3. I've had very good results having models from one provider critique plans and code written by models from another provider. Can't do that without subs to at least two providers

What intrigues me more is how are they not done building whatever they were building? LLMs have been jaw-droppingly good for a good 6 months now.

The people spending thousands on LLM bills a month should've been done building significant chunks of Microsoft's or Google's software portfolio by now.

I guess what I mean is how is producing code still the limiting factor in whatever they're doing?

I produce a firehose of personal software and I still cannot max out a $20 plan.

Scope creep? When you give a mouse a cookie and all that...

If it's $1000 per month to make $0 because you haven't found PMF, less obviously a good deal unless you have a pathway to revenue?

I've been dropping the equivalent of a car payment for three months and it's been worth every single penny; even if I never actually sell a product.

https://github.com/ByteTerrace/Puck

It was originally a small project in scope, just a basic GameBoy emulator, a shader VM, and some Azure integrations. Just a bunch of things I wanted to dabble with in order to learn and have fun. Over the past few months it has grown immensely into something resembling an operating system more than a toy engine. There's even a custom DSL for authoring worlds that abstracts over the state/rendering models; it's completely ridiculous!

Will I make any money off the thing? Probably not, don't care. The things I've learned along the way have been worth way more than the amount of money I burned on a few monthly subscriptions. Learning how to interact with agents and treat them like a team was worth it alone; the product I have now is simply a nice bonus.

I think you can just become slightly careless and off it goes. I just have a £20/mo ChatGPT and I'm getting it to do some HomeLab Kubernetes YAML for me that I have no desire to hand-type. I asked it a question and it zoomed off looking at all sorts of websites, running various git commands, doing anything except asking me for an answer to a question.

It ran through about 60% of my 5-hour allocation in those few minutes, which was an unheard-of rate for me before that, and I can imagine if I had zillions of tokens available I just wouldn't have noticed. As it is, my AGENTS.md now exists and is fairly stern.

The highest is Pro 100 and it’s useless for codex.

> As a thank you for being an existing subscriber and to help you continue to build your most ambitious ideas and transition, we’re also giving you a one-time grant of 62,500 usage credits (a $2,500 USD value). You’ll see these appear in Settings > Usage. They will expire on December 31, 2026, and use of the credits is governed by the Service Credit Terms.

I have no idea how much 62,500 usage credits are. Given the value they suggest I feel it's like one extra week of usage.

With how many resets they've been giving out lately, if you consistently max things out I would say this is a reasonable guess. You can ask your clanker via your favorite harness (the usage credits are equivalent to API pricing):

> Based on my session files and official OpenAI API pricing, check how much I would have spent in the past few months if I switched to the OpenAI API.

Since February

I've been paying for the PRO200 plan since day one, and I've never used it that much because Claude Code was and still is better at coding. I didn't even complain when they started nerfing their models.

I love OpenAI, but they should do better on such important decisions, knowing that all other companies are waiting for such missteps.

OpenAI must have statistics on what percentage of people with Pro accounts don't maximize the usage in a given. Even if it a very small percentage that is still income for providing nothing.

When you think about it, it's kind of ridiculous that the usage credit is use-it-or-lose it. It really should roll over.