I think the whole nature of agentic programming is variable reward: kick off a bunch of prompts, obsessively check “which one nailed it??” Or at least I get a dopamine hit when the agents notifies me that it’s done with its turn.
I don't really know if really fits the description, but the whole effort (`\effort`) gate is pretty on point. I mean, you can go up to Ultracode which takes over the whole bottom portion of the terminal with a shiny glow, Ultracode for sure give a lot of "up the ante" vibes...
Timers. Such as these five hour and seven day limits.
If you see any sort of timer anywhere, you are in a Skinner's box simulator. You are being subjected to positive and negative reinforcement via the reward schedule. This is habit forming.
It's definitely addictive and it does create problems.
The timer only starts when we message a model, and we always want the timer to be counting down. So if a timer resets at 3 AM, we are incentivized wake up at 3 AM and start something just to get the timer running. Mercifully, unlike actual video games, automating this isn't cheating.
The existence of timer resets also incentivizes us to spend our usage as much as possible. No way to know when Tibo's gonna push his reset button. Ideally, we want to be at zero percent remaining when he does. Any usage left before reset is usage that was left on the table. The new 5h limits will make this much more difficult, and this is probably by design.
Heh, it's just like my mobile gaming days... Deal with this long enough and eventually you can see the game mechanics like Neo sees the Matrix. Never thought I'd fall into this rabbit hole again.
I disagree about the "what did they build" angle though. These subscriptions are providing me with truly absurd amounts of value. Gacha games deserve to be mercilessly botted to death, but AI is world changing technology.
It's a little more like App Store "gems" imo, with some clarity and honesty removed. Pay for "usage" that disappears arbitrarily: coded that bug fix for you, 3% usage consumed. Would you like to upgrade your plan to pro for "more" usage or max for "much more" usage? Buy pro max for "most" usage and access to our latest model, until we revoke it 15 minutes from now.
>> During this period, the company also reset users’ weekly usage early on several occasions to celebrate milestones, including adding another million active users across the two (now unified) platforms.
> Such a casino vibe.
No no no! It means we're all in it together! Feel the parasocial vibes!
I used to think it was pretty great but now it just means you can't plan your usage across the week.
It's a slightly entitled complaint but when you 'waste' 80% of your remaining usage because you get a reset you weren't expecting, there's a ping of disappointment there.
If you don't like it just ignore it. You're just trying to game the system and extract the absolute most you can and you're upset that this thwarts your game and doesn't benefit you. People like you are why we can't have nice things
I've seen a lot of people who seem to think it's inevitable that they'll get resets, making them more likely to over-spend early because they'll get a reset anyhow.
It could leave them out of availability later when they need it.
I think there can be many complaints about large scale AI providers but given that large companies are now pushing for local models it seems that we are luckily avoiding the situation where AI is a utility
They now give out resets you can choose when to use, but with a "best before" date. With Anthropic you don't even know what model you'll have access to next week. A few months ago I wanted to try a subscription to Gemini but I couldn't figure out how to give them money. The other one thinks he's Mecha Hitler.
To an extent I get it, this is not like a normal SaaS where you just buy seats, but their commercial offerings, especially B2C ones, feel like complete amateur hour. If they were selling literally anything else I'd have ran for the hills a long time ago.
Yeah. I switched to codex 1.5 months ago to try it, and for some reason the "vibe" kept me here. It was really the unpredictable rewards from the constant resets though. I am not convinced it's a better product or that sol + any combination of luna is better than Claude.
> This is necessary as (a) the 5h limit allows us to smoothen the load on our compute, allowing to keep the plan generous in terms of weekly usage
Seems fair, but then maybe the 5h usage should also drain slower during off-peak hours? Perhaps it already does.
> and (b) users on the Plus plan are relatively casual and new users, but then also just accidentally eat through their whole weeks usage and then are confused, making it not a great experience.
Wouldn't this one be fixed by making the 5h limit a guardrail you can opt out of? If so, then this doesn't work as justification for a mandatory limit.
Also, I want to be able to eat through my whole week's usage at once. What if I was busy all week and want to work on my personal project(s) Saturday or Sunday.
That is a pathological use case for OpenAI, and precisely why they have instituted these limits. Part of the tradeoff of doing your work on someone else’s computer is they get to set the rules about when you use it.
What was described is not pathological. It doesn't involve disease or a disturbed mental state. No part is paradoxical, counterintuitive, or involves unexpected results.
It is normal and healthy behavior that is produced from within a rational mental state. The complaint about scheduling is also normal and rational.
The closest thing to "pathological" that I'm seeing here in these comments is trendy misuse of the term.
It is pathological for their utilization, as in “worst-case behavior.” Like how a program that makes and frees millions of differently-sized heap allocations is pathological for memory fragmentation.
I’ve been hit with limits at 4am Eastern time when I’d expect for not a lot to be going on. I always thought the quotas would be used best for shaping demand curves, but as you said — there are no off-peak hours anymore.
Given that it's not possible to buy the USD 100,- or USD 200,- subscription on a team/corporate plan this effectively introduces the 5h limits again for all corporate users.
At this point the 40 USD/120 USD Cursor Team subscription suddenly becomes a lot more attractive. You'll get more generous limits and a single monthly limit.
I'm curious about why you see ads, I didn't think anyone on HN does, nowadays. Given how easy it is to install uBlock, is it a conscious decision to browse adfully?
Some of us don’t use an adblocker because we are aware that ads are how most websites get paid for on the internet. I pay for sites I care to support, but for something like LinkedIn, I’d rather have their monetization scheme push me away from any usage than block the ads and make the site slightly more palatable.
LinkedIn has a couple of useful features: Getting in touch with former coworkers; seeing who might know the hiring manager for a position you're applying for; uh... that's about it.
The rest is garbage but I guess is how they make their money.
Why not? I owe my life to ads. Thank you Google AdSense and Yahoo! Ads for the hundreds of thousands of dollars you paid for 16 year old Rob to slap them on his MySpace layout website in 2006 so Timmy could put a glitter graphic of Stone Cold Steve Austin on his banner.
probably less - I sometimes manage to burn like most of my 5h credits on Claude 5x Max plan in 2h.
The annoying things is that you gave bigger tasks and is such session finished maybe 70% and after 5h pool resets cache will expire so that if you "please continue" it will wipe out again a lot of your 5h pool)
This sucks, but really the value of the $20 plan has been insane. I routinely get weeks of work out of the weekly limit, so yeah I'll just move up plans if needed.
Tighter limits, forcing people to higher plans. Most will go as they are addicted and dependent on those tools. They need to use to finish the project they started as no human will touch that pile of code. If you didn't saw this coming, as it has been the standard business strategy from all kinds of SaaS in the last years, you haven't been paying attention.
There are solid, cheap options like xAI, z.AI, MiniMax, Kimi, DeepSeek, and Alibaba. Heck, I can run a coding agent on my 32GB MacBook Pro M1 right now and it's actually pretty good (just slow). No one is forced to keep paying OpenAI and Anthropic. That's their major strategic risk: distilled models are far cheaper to build and run and they're only a few months behind.
Imagine being a student and using codex for some night studies and hit the limit at the wrong time saying let's continue after exams, still we have normal chat available.
Maybe for projects also but we still have no limits in normal chat section but that ain't codex for working on our projects.
Personally I have been using Deepseek V4 flash and it is enough for my needs. You can use OpenCode Zen/Go and also get access to many other models, including OpenAI ones. Writing purpose specific agents can make cheaper models close the intelligence gap with the most expensive ones
How is that the case? ChatGPT not only gives a lot of Codex/Agentic usage, once you've run out of that you still have a massive quota of Chat usage that's entirely separate. You can still use plug-ins with Github, Google Drive, etc... using the Chat usage as well.
There are many providers and harnesses. You could go almost 90% of the frontier models for the fraction of the price. You have 3 more free days of ox alpha.
Geofencing has been quite good at creating feeling of being ripped off in the developed countries - from which the bulk of the companies revenue come from - that companies seems to be dropping it lately.
Anyway - China's advances seems to have caught the big 2 unprepared, so price war seems to be going on.
And there are Chinese resellers for frontier models if you are desperate for high end frontier tokens.
In the last 30 days my LLM spend, if paid at API rates, would be $19,431 (and that doesn't track online chats or my Github Action reviews). Now I wish I was making $200K/mo but I'm not, no where close. I got all that for $220 on subscriptions.
Yes, LLMs are awesome and let me do things faster but I would be using them far less if my only option was the API. I used coding agents prior to subscriptions (Aider) and spent <$200 total before abandoning it due to cost. The results were good, but not worth the price for me.
If OpenAI were actually serious about capturing marketshare they'd release an unlimited subscription plan. Rate limit users if you need to, but let people use their agents 24/7. Even at $1000/month most would take it.
Is there a world where we look back at how AI usage is charged today and we equate it with how we had minutes on AOL and how absurd it seems looking back?
Absolutely. LLM providers are likely to be the next telecom providers. Low and steady income, translating to safe dividend income for investors. Most of these companies have no moat, since most LLM implementation strategies more or less converge to the same result.
I think without a doubt that will be the case, unless the trend of compute getting better over the last 60 years suddenly stops. It should become less tough to run a local model and our devices should become more powerful.
However I think there is still a significant runway for these models to scale, so there will always be some sort of offering from providers. I can't imagine that our current use of the context window will be how that looks in a handful of years.
Absolutely. Tokens will flow like electricity, in the instances when you aren't just paying for electricity to run models locally. The token providers (well, at least the frontier labs) will fight it tooth-and-nail with the example of telecoms being "how to become a commodity" but as long as open models continue they will have no choice.
API costs are simply too high. If you go wild you can easily burn $3-5k a day. If you really hammer it (ie going full parallel agentic) you can exceed $10-20k. And this isn't even going crazy. Those costs are simply too high to make any sense.
I think you're underestimating the "more expensive". It's not a 30% discount or similar. API is at least 10 times more expensive (own experience), ~40x as expensive if [1] is accurate.
At that price difference, lots of things that are viable to do in whatever harness the subscription allows are simply impractical via API, you'd be spending a whole developer's salary just for the benefit of using a different harness.
I tried for example Deepseek V4 Pro for a bit (via API pricing), but it's _more expensive_ than Sol via subscription, even though $1.7/Mtok is a lot less than 20$/Mtok.
> API is at least 10 times more expensive (own experience),
More. On CC I'm currently burning ~$5k a day (according to claudes own metrics) and I use up the weekly quota in about 3-4 days. Meaning I'm eating $20k worth of tokens for a $200 sub.
Current session I have running in the background, spinning up tests mostly
> At that price difference, lots of things that are viable to do in whatever harness the subscription allows are simply impractical via API, you'd be spending a whole developer's salary just for the benefit of using a different harness.
Could you elaborate on that?
Off topic:
You did the ISBN visualization thing! Awesome!
Haha. I've been pretty happy using similar limits on Claude precisely for this. It provides a really nice work:life balance without that nagging feeling that I should really get a few more hours in.
"During this period, the company also reset users’ weekly usage early on several occasions to celebrate milestones"
Such a casino vibe.
Hey I work in the casino industry and we have regulations!
Both OpenAI and Anthropic use a lot of casino/slot machine mechanics in their TUIs and Apps so likely intentional.
Do you have any examples of this? I am not sure if I understand what you mean
I think the whole nature of agentic programming is variable reward: kick off a bunch of prompts, obsessively check “which one nailed it??” Or at least I get a dopamine hit when the agents notifies me that it’s done with its turn.
I don't really know if really fits the description, but the whole effort (`\effort`) gate is pretty on point. I mean, you can go up to Ultracode which takes over the whole bottom portion of the terminal with a shiny glow, Ultracode for sure give a lot of "up the ante" vibes...
Timers. Such as these five hour and seven day limits.
If you see any sort of timer anywhere, you are in a Skinner's box simulator. You are being subjected to positive and negative reinforcement via the reward schedule. This is habit forming.
explains the junkie behavior from the boosters. up all night just one more prompt and what did they build? well it's almost done
it's like nitrous oxide, you're sure to have that epiphany with just one more pull
It's definitely addictive and it does create problems.
The timer only starts when we message a model, and we always want the timer to be counting down. So if a timer resets at 3 AM, we are incentivized wake up at 3 AM and start something just to get the timer running. Mercifully, unlike actual video games, automating this isn't cheating.
The existence of timer resets also incentivizes us to spend our usage as much as possible. No way to know when Tibo's gonna push his reset button. Ideally, we want to be at zero percent remaining when he does. Any usage left before reset is usage that was left on the table. The new 5h limits will make this much more difficult, and this is probably by design.
Heh, it's just like my mobile gaming days... Deal with this long enough and eventually you can see the game mechanics like Neo sees the Matrix. Never thought I'd fall into this rabbit hole again.
I disagree about the "what did they build" angle though. These subscriptions are providing me with truly absurd amounts of value. Gacha games deserve to be mercilessly botted to death, but AI is world changing technology.
It's a little more like App Store "gems" imo, with some clarity and honesty removed. Pay for "usage" that disappears arbitrarily: coded that bug fix for you, 3% usage consumed. Would you like to upgrade your plan to pro for "more" usage or max for "much more" usage? Buy pro max for "most" usage and access to our latest model, until we revoke it 15 minutes from now.
When there’s billions involved everything is intentional
It's the gacha of software development.
>> During this period, the company also reset users’ weekly usage early on several occasions to celebrate milestones, including adding another million active users across the two (now unified) platforms.
> Such a casino vibe.
No no no! It means we're all in it together! Feel the parasocial vibes!
also it sounds like users benefit but they mostly don't. resets mostly benefitted OpenAI and only gave users few additionals hours.
I think it's great that oai resets users limits. What is this complaint?
Free drinks? Keep customers happy? Reward your heaviest users? Casino vibes
I used to think it was pretty great but now it just means you can't plan your usage across the week.
It's a slightly entitled complaint but when you 'waste' 80% of your remaining usage because you get a reset you weren't expecting, there's a ping of disappointment there.
If you don't like it just ignore it. You're just trying to game the system and extract the absolute most you can and you're upset that this thwarts your game and doesn't benefit you. People like you are why we can't have nice things
I've seen a lot of people who seem to think it's inevitable that they'll get resets, making them more likely to over-spend early because they'll get a reset anyhow.
It could leave them out of availability later when they need it.
Well, you know what they say. The first one's free.
When did hn turn into reddit?
I think there can be many complaints about large scale AI providers but given that large companies are now pushing for local models it seems that we are luckily avoiding the situation where AI is a utility
They now give out resets you can choose when to use, but with a "best before" date. With Anthropic you don't even know what model you'll have access to next week. A few months ago I wanted to try a subscription to Gemini but I couldn't figure out how to give them money. The other one thinks he's Mecha Hitler.
To an extent I get it, this is not like a normal SaaS where you just buy seats, but their commercial offerings, especially B2C ones, feel like complete amateur hour. If they were selling literally anything else I'd have ran for the hills a long time ago.
The vibe was people goofing around on Twitter. There's one guy making the call and the reset is kind of a meme.
Yeah. I switched to codex 1.5 months ago to try it, and for some reason the "vibe" kept me here. It was really the unpredictable rewards from the constant resets though. I am not convinced it's a better product or that sol + any combination of luna is better than Claude.
I'll be switching back.
> This is necessary as (a) the 5h limit allows us to smoothen the load on our compute, allowing to keep the plan generous in terms of weekly usage
Seems fair, but then maybe the 5h usage should also drain slower during off-peak hours? Perhaps it already does.
> and (b) users on the Plus plan are relatively casual and new users, but then also just accidentally eat through their whole weeks usage and then are confused, making it not a great experience.
Wouldn't this one be fixed by making the 5h limit a guardrail you can opt out of? If so, then this doesn't work as justification for a mandatory limit.
It feels like daily quests in MMOs, trying to force user engagement via cultivating addiction-style behavior patterns
Come back tomorrow so you can raise your faction reputations another 0.00001%!
Also, I want to be able to eat through my whole week's usage at once. What if I was busy all week and want to work on my personal project(s) Saturday or Sunday.
That is a pathological use case for OpenAI, and precisely why they have instituted these limits. Part of the tradeoff of doing your work on someone else’s computer is they get to set the rules about when you use it.
How is that pathological? It sounds like a completely normal behavior.
What was described is not pathological. It doesn't involve disease or a disturbed mental state. No part is paradoxical, counterintuitive, or involves unexpected results.
It is normal and healthy behavior that is produced from within a rational mental state. The complaint about scheduling is also normal and rational.
The closest thing to "pathological" that I'm seeing here in these comments is trendy misuse of the term.
It is pathological for their utilization, as in “worst-case behavior.” Like how a program that makes and frees millions of differently-sized heap allocations is pathological for memory fragmentation.
There are no off-peak hours. I have received the model overloaded error that cancels my current goal during the early hours of the morning.
> early hours of the morning
which TZ?
> There are no off-peak hours.
Probably US west coast hours, but there's no public stats
2AM CST
I’ve been hit with limits at 4am Eastern time when I’d expect for not a lot to be going on. I always thought the quotas would be used best for shaping demand curves, but as you said — there are no off-peak hours anymore.
Given that it's not possible to buy the USD 100,- or USD 200,- subscription on a team/corporate plan this effectively introduces the 5h limits again for all corporate users.
At this point the 40 USD/120 USD Cursor Team subscription suddenly becomes a lot more attractive. You'll get more generous limits and a single monthly limit.
Premium seats are coming soon.
https://openai.com/index/premium-seats-chatgpt-business/
Based on the ads flooding me on LinkedIn, Facebook, and anywhere else I see ads, Cursor is really going after this market right now.
I'm curious about why you see ads, I didn't think anyone on HN does, nowadays. Given how easy it is to install uBlock, is it a conscious decision to browse adfully?
Some of us don’t use an adblocker because we are aware that ads are how most websites get paid for on the internet. I pay for sites I care to support, but for something like LinkedIn, I’d rather have their monetization scheme push me away from any usage than block the ads and make the site slightly more palatable.
No, ads mean the customer is the advertiser, not me, the user, and align the business' interest with the advertiser's, against my own.
I'd rather the site asked me to pay directly, or not at all. I will not see any ads I can avoid.
I don't think that many people here voluntarily go on Facebook either
I’m still shocked to hear that people go on LinkedIn
update it once a year with the resume. sometimes a mid-year connect to coworkers or respond to someone I know who poked me.
more than once a quarter? nah. would also delete it if I could
LinkedIn has a couple of useful features: Getting in touch with former coworkers; seeing who might know the hiring manager for a position you're applying for; uh... that's about it.
The rest is garbage but I guess is how they make their money.
yeah do yall go on linkedin for anything legitimate??
Why not? I owe my life to ads. Thank you Google AdSense and Yahoo! Ads for the hundreds of thousands of dollars you paid for 16 year old Rob to slap them on his MySpace layout website in 2006 so Timmy could put a glitter graphic of Stone Cold Steve Austin on his banner.
> Given that it's not possible to buy the USD 100,- or USD 200,- subscription on a team/corporate plan
https://openai.com/index/premium-seats-chatgpt-business/
So we must move to 5 hour working days. Thanks OpenAI!
probably less - I sometimes manage to burn like most of my 5h credits on Claude 5x Max plan in 2h.
The annoying things is that you gave bigger tasks and is such session finished maybe 70% and after 5h pool resets cache will expire so that if you "please continue" it will wipe out again a lot of your 5h pool)
Lets normalize working 5 hours a day. :)
I really don't mind blowing my weekly usage in one day, but getting my tasks stopped and waiting for 5h cap to finish - is just annoying.
https://codex-resets.com/
The reset cadence is absurd.
This sucks, but really the value of the $20 plan has been insane. I routinely get weeks of work out of the weekly limit, so yeah I'll just move up plans if needed.
It will get worse!
Tighter limits, forcing people to higher plans. Most will go as they are addicted and dependent on those tools. They need to use to finish the project they started as no human will touch that pile of code. If you didn't saw this coming, as it has been the standard business strategy from all kinds of SaaS in the last years, you haven't been paying attention.
There are solid, cheap options like xAI, z.AI, MiniMax, Kimi, DeepSeek, and Alibaba. Heck, I can run a coding agent on my 32GB MacBook Pro M1 right now and it's actually pretty good (just slow). No one is forced to keep paying OpenAI and Anthropic. That's their major strategic risk: distilled models are far cheaper to build and run and they're only a few months behind.
Imagine being a student and using codex for some night studies and hit the limit at the wrong time saying let's continue after exams, still we have normal chat available. Maybe for projects also but we still have no limits in normal chat section but that ain't codex for working on our projects.
This is such a bummer for those of us living in third world countries. I’m not a casual user, I just can’t afford to pay more.
Personally I have been using Deepseek V4 flash and it is enough for my needs. You can use OpenCode Zen/Go and also get access to many other models, including OpenAI ones. Writing purpose specific agents can make cheaper models close the intelligence gap with the most expensive ones
Aren’t you better using other services than OpenAI if the price matters to you?
I’ve tried Claude Code and recently switched to Codex. I’ve found that $20 on Codex gets me 4–5 times more usage than on Claude Code.
I could burn through my Opus usage in 1 hour whereas with Sol, I can go 3–4 days.
Oolama? That plan is more than generous for $20.
How is that the case? ChatGPT not only gives a lot of Codex/Agentic usage, once you've run out of that you still have a massive quota of Chat usage that's entirely separate. You can still use plug-ins with Github, Google Drive, etc... using the Chat usage as well.
From my experience:
* OpenAI was ~$170 per week in value
* Claude was around $500 per week in value.
* Ollama Cloud is giving me about ~$160 per week in value.
So right now (thing constantly change in the AI world), its not more generous then Claude.
You might be thinking of API pricing? The subscriptions give you a lot.
Oh right, thanks for catching that
There are many providers and harnesses. You could go almost 90% of the frontier models for the fraction of the price. You have 3 more free days of ox alpha.
Geofencing has been quite good at creating feeling of being ripped off in the developed countries - from which the bulk of the companies revenue come from - that companies seems to be dropping it lately.
Anyway - China's advances seems to have caught the big 2 unprepared, so price war seems to be going on.
And there are Chinese resellers for frontier models if you are desperate for high end frontier tokens.
I left Claude because of this. I guess now I need to move on from OpenAI also
But where will you go?
I really don’t get why people use the native chat apps when you can use all models via their APIs with no restrictions.
It’s more expensive, yes, but I assume most people here use it for something that is worth spending money on?
From my rough understanding, tokens are about five to ten times more expensive than the prepaid plans
I get that it’s not nothing, but from personal experience it makes me just so much faster.
I’m basically replicating myself multiple times for a ~10% salary surcharge.
In my mind it’s impossible to not see it that way, unless you just want to capture the subsidy surplus for free.
In the last 30 days my LLM spend, if paid at API rates, would be $19,431 (and that doesn't track online chats or my Github Action reviews). Now I wish I was making $200K/mo but I'm not, no where close. I got all that for $220 on subscriptions.
Yes, LLMs are awesome and let me do things faster but I would be using them far less if my only option was the API. I used coding agents prior to subscriptions (Aider) and spent <$200 total before abandoning it due to cost. The results were good, but not worth the price for me.
It seems you’ve stumbled into the truth: a lot of things are only worth throwing LLMs at as long as the LLMs are cheap.
Chinese models.
I see this as evidence that last weeks token burn rate wasn’t by accident or due to some “bug.”
I don’t mind the 5 hour window but this is a reminder to not hinge your workflows on any one tool.
I’m not a user myself but find it not great for your customers to announce such a change for the next day. Cannot they announce a few days in advance?
If OpenAI were actually serious about capturing marketshare they'd release an unlimited subscription plan. Rate limit users if you need to, but let people use their agents 24/7. Even at $1000/month most would take it.
Is there a world where we look back at how AI usage is charged today and we equate it with how we had minutes on AOL and how absurd it seems looking back?
Absolutely. LLM providers are likely to be the next telecom providers. Low and steady income, translating to safe dividend income for investors. Most of these companies have no moat, since most LLM implementation strategies more or less converge to the same result.
I think without a doubt that will be the case, unless the trend of compute getting better over the last 60 years suddenly stops. It should become less tough to run a local model and our devices should become more powerful.
However I think there is still a significant runway for these models to scale, so there will always be some sort of offering from providers. I can't imagine that our current use of the context window will be how that looks in a handful of years.
Absolutely. Tokens will flow like electricity, in the instances when you aren't just paying for electricity to run models locally. The token providers (well, at least the frontier labs) will fight it tooth-and-nail with the example of telecoms being "how to become a commodity" but as long as open models continue they will have no choice.
Any other outcome would be pretty terrible.
Not when you consider the sheer amount of compute to calculate a single token. Today, at least.
Maybe if we have some breakthrough in how to get equivalent ai capacity out of less compute it’ll seem crazy in retrospect.
But the magnitude of flops per token on SOTA models is mind boggling huge.
AOL minutes seem antiquated, but not absurd.
Eventually.
Maybe I should go back to ChatGPT again and copy-paste code.
Stupid question: Why are y'all not using the API? It’s more expensive, but you can do what you want.
> It's more expensive
There's your answer
Man, that was my instinct too, but I’m not sure.
I don’t think everyone here is using Claude/GPT/… to create indie games?
If you’re using it for work then waiting another week/day/hour is a business expense, effectively.
Maybe it's like this guy I know who purchased a new-ish Bugatti but all the service is done by a local mechanic, during his time off, in a barn.
API costs are simply too high. If you go wild you can easily burn $3-5k a day. If you really hammer it (ie going full parallel agentic) you can exceed $10-20k. And this isn't even going crazy. Those costs are simply too high to make any sense.
I think you're underestimating the "more expensive". It's not a 30% discount or similar. API is at least 10 times more expensive (own experience), ~40x as expensive if [1] is accurate.
At that price difference, lots of things that are viable to do in whatever harness the subscription allows are simply impractical via API, you'd be spending a whole developer's salary just for the benefit of using a different harness.
I tried for example Deepseek V4 Pro for a bit (via API pricing), but it's _more expensive_ than Sol via subscription, even though $1.7/Mtok is a lot less than 20$/Mtok.
[1]: https://tibotattle.com/
> API is at least 10 times more expensive (own experience),
More. On CC I'm currently burning ~$5k a day (according to claudes own metrics) and I use up the weekly quota in about 3-4 days. Meaning I'm eating $20k worth of tokens for a $200 sub.
Current session I have running in the background, spinning up tests mostly
Total cost: $660.33
Thank you.
> At that price difference, lots of things that are viable to do in whatever harness the subscription allows are simply impractical via API, you'd be spending a whole developer's salary just for the benefit of using a different harness.
Could you elaborate on that?
Off topic: You did the ISBN visualization thing! Awesome!
I can understand Plus to some extent but why would Work users be considered casual or new?
No point in using /fast mode if you are gonna get stopped after 5 hours.
Fast mode doesn't use subscription usage (IIRC) so it's not going to be limited to 5hr.
I say its good because you can atleast have a break in between sessions
Haha. I've been pretty happy using similar limits on Claude precisely for this. It provides a really nice work:life balance without that nagging feeling that I should really get a few more hours in.
Tragic day in AI.
The most transformative technology ever created! Brought to you in 5 hour increments..... maybe
They are conditioning us for the soon-to-come electricity brown-outs. </only half joking>
Unlocking the weekly limit coming soon for $9.99.
Honestly didn't expect that
Please don’t! Don’t limit my daily usage
Imagine paying for this shit..