I think the difference between the Anthropic token maximization approach (vibe code all the things!) and OpenAI's focus on efficiency, terseness and token reduction are going to be the defining features of who wins the long-term race.
My money is on the more efficient solution. Even if Anthropic can win some benchmarks by using 3x tokens over 3x time, it is a terrible base to build toward the future. Users are no longer willing to wait exponentially long for linear improvements. And as we see from some of the Chinese models, they can quickly distill frontier models with the tax of being slower and more token-guzzling, while retaining most of the quality. The real differentiators are becoming speed and efficiency, which translate to cost and user velocity more than incremental capability improvements.
It's great that frontier models can solve complex math equations, but bread and butter LLM usage (where the money is made) has already shifted from "I need the best always" to "what solves my day-to-day problems quickly and consistently". Fable usage as a percentage is flat-lining. We are already at the point where output quality is negligible. What wins going forward is cost, speed, consistency and the compounding effects of "softer" improvements to the harness.
You've articulated what I found unsettling about Boris' pov that "coding is a solved problem". I listened to a few of his talks and was instinctively off-put by that sentiment. I figured a fellow programmer would understand and speak on the nuances.
Granted it did make me think about my biases and to lean into more future facing inevitabilities. But you've nailed it, for Boris and Anthropic, they are betting that coding is a solved problem in the sense that any person can one-shot any random idea and the output will be in some abstract sense "good". And then at what cost and toward what end?
I really would like to see OpenAI’s focus on efficiency but everytime I use Codex, it wastes tokens like there’s no tomorrow, hitting week limit in a day, where I’m able to use Claude just fine. Maybe it’s based on the codebase, I don’t know, but I have better results with Claude than Codex.
it'll all be centralized models justifying their capability against local inference in, maybe 3 years or less
once we have a bit more memory fab capacity and the insane bottomless investment in AI giants realizes there is a bottom, local hardware will catch up with model performance to the extent that centralized inference will be downgraded to special cases or for orgs that find it cheaper than buying expensive hardware
but really most power users are going to have 1 TB unified RAM and local models that will do well enough
for light office use you can still have a cheap laptop and a claude subscription
Oh no, I use the top shelf models every day and I really think they have a lot of room to improve in pretty much every regard. I suspect Fable usage is flatlining because it's not that good comparitively and way too expensive.
For anyone on the fence, I was a hardcore CC user since it was released. I made the full switch to 5.6 sol and Codex about a month ago.
It's the better experience. The limits are way higher (I almost never burn through my $200/m plan), and the output is better than Opus 4.8 (Opus 5 is completely unusable for me).
It's hard to put a number on it, but even accounting for all the time in meetings, talking with stakeholders and developers etc I'm over 10% more productive overall. I earn substantially more than $2000/month, the ROI is there
It's only expensive compared to the currently very strong offering from OpenAI. Or other models - my hobby projects are all on DeepSeek
I switched when 4.7 dropped, after over a year on the service. It was an immediate improvement - not even 1 hour before Codex was giving me better results with less effort.
Stick with xhigh rather than turbo for most tasks though, now that the Age of Abundant Resets seems to be well and truly over for good (and 5h limits are supposedly returning soon).
During this period however they’ve released models which lots of people report to be significantly more chatty. Framing it purely as a promotion ending feels like you’re giving them a bit too much credit. I doubt it’s a coincidence they started this promotion a week before the release of 4.8.
I think running a model unsupervised in anything above medium is a sucker move to burn more tokens. The high effort models can be great in limited context, but unsupervised they too often end up navel gazing. High doesn't always mean smarter, but it always burns more tokens. Medium and low seem to be decent for day to day tasks.
I am not sure the supervision is the problem, it's more the way Claude is acting in general, taking initiative that could burn thousands of token in a second. Just today I ask it a small fix on a newly pulled project, it wrote 5 lines of code but then immediately started to run the full test suites, which itself runs in docker, with several images that I didn't pull yet so it started to download everything and by the time I click stop, it already filled the context windows with dozen of hundreds of useless lines...
I find it very hard to fins the sweet spot in the prompt between "do not assume, do not take initiatives" and "inform me if there is a simpler way to do what I am asking or if there are some downsides I may not have thought about".
That’s my experience too, going off the deep end with CoT doesn’t meaningfully improve output on a lot of tasks but it does incinerate tokens and take longer
I found the waffle reduced my overall usage. It doesn't matter much if it's smarter if it also needs more babysitting. It's quite good at making something quite advanced that is tangentially related to what you actually wanted.
As soon as it gets annoying enough to switch to another provider your Claude code tokens drop right off.
They've been really struggling all of 2026. All of these "limited time" promos, just to have less usage than codex, and the shenanigans around "peak hour" reduction earlier in the year.
I got a max account during this promotion and it's been very fun but I'm a little burned out and am kind of looking forward to going back and tinkering with game engines without the help of an LLM for code gen (will still use it for documentation questions but I can do that with the free tier)
Ah, forgot about that. I was questioning if I should switch my big subscription to ChatGPT (was on the 20$ Open AI and 100$ Claude), seems like a no brainer now
I've been watching this to see if they "extend" it again, let it end, or make it permanent. I'm regularly hitting 90-100% on my $200/mo sub and will switch to Codex in a heartbeat if they drop these limits. This constant uncertainty is really annoying.
Why, it seems like the supply is limited by the ability to provide, and the demand is limited by market alienation.
They raise their prices too high, a lot of customers will still buy but be unhappy about it, it's burning goodwill for money. If they drop their prices too low they have overwhelming demand.
They could be making money hand over fist for all we know. We don't really know how much compute is being used, nor how much it cost them.
To anyone like me that switches between Codex and CC based on limits, but prefers CC to Codex, PI has been really great to migrate over. It can customize itself very easily, I was expecting to have to clone the repo or whatever, but no you can just tell it "add a manual mode where you present all changes as diffs in VSCode, letting me edit them or save to accept" and it will do it.
I wish I could use my Anthropic sub with it but I heard you get banned, but at least you can use it with any other subscription or model.
I suspect the net result will be that more Pro $100 subscribers upgrade to $200/mo. If you explicitly do simpler stuff with Opus, you generally have enough Fable time for most programming and planning tasks.
No. Here's a data point: I tried it for a few days. While it worked fine I didn't see any benefit to switching since it costs the same ($100) and I've spent a year getting up to speed on Claude usage and quirks.
If Anthropic doesn't figure out why OpenAI is eating their lunch, they're in some trouble. I'm guessing they already know why, but aren't able to accept that they won't somehow come out #1. Or in fact they really believe they will still come out #1. Delusion?
People take much more time off during the summer in the USA than during other times of year. It’s somewhat regional, but it’s been true in all the place I have lived. Probably July is the most common month though because in some regions schools start sometime in August. It’s not uncommon for people to take 1-2w off during that period. More than 2w often requires extra approvals, so is less common.
I saw this announcement when it came out and completely forgot about it. That explains why Claude Code felt so surprisingly generous these past few months.
I forgot this was in effect. I’ve been hitting my weekly rate limits two to three days in because of needing to steer Slopus 5 with Fable. Absurd that they’re being cut further in the face of steep competition from open weights models.
I actually use claude these days mostly for mission critical tasks and code reviews. Everything else I use Sol. I have $20/month plans on both. For non-coding I also use Gemini on a $20/plan. So far I haven't had any issues but we'll see how things go. I have Qwen3.8-27B installed locally but until I upgrade my mac it's not for day to day stuff.
As we speak, Claude is down. I was already complaining about how extremely slow and dumb it has gotten since Opus 5. I guess this will be the last nail in the coffin for Claude. At least for me.
The 'tokenmaxxing' and the extreme gambling fiesta with Anthropic's latest Claude Code slot machine engine called 'Fable 5' which has given many subsidies and free spins of the wheel and cheaper tokens are all incompatible with the desires of their future owner: Wall Street.
If Wall Street sees a single outage or a tiny drop in usage, they won't be happy and will pressure Anthropic to take away the free tokens.
Better to reduce the limits now rather than to wait until Wall St. tells them to just to avoid a stock punishment.
Claude Opus and Fable are so bad compared to GPT-5.6-Sol it's ridiculous, their desktop client is worse, and the value is worse because OpenAI has been spamming discounts. They better get their shit together at Anthropic!
Edit: If you're mad I'm light on the details I have provided some in the replies.
gpt-5.6-sol uses fewer tokens and OpenAI has been more generous with usage, but quality-wise I wouldn't say they are inferior to each other.
I have a Claude Code and OpenAI subscription so that I can use Opus/Fable/gpt-5.6 as I please, and the models are often catching things the other models missed. So much that I would significantly weaken my workflow if I dropped one subscription.
My best workflow at the moment is to create the initial plan with Fable (before review/revise-cycling with other models). From my own testing it seems slightly better at arriving at high-level ideal solutions after sweeping the whole project, projecting future needs, then coming up with good trade-offs like "by construction" correctness.
While mostly subjective, maybe the closest objectivity I have here is noticing fewer revision cycles needed with Fable-initialized plans.
For personal projects, I defer almost everything day-to-day to the models, so it gives me a good feel for what the models are good at.
Even then, it's kind of a wash these days between the sota models, and we're talking about maybe a 10% performance difference or something. But every once in a while there's the experience of one model spinning its wheels on a bug/repro/issue while another model comes in and one-shots the solution.
That make sense. Let me ask you this
When the project reaches a level of complexity . Do you simply reach for better models or do you reengineer it or does the project scope stop at the egdge of the model's capabilities.
I am asking because in my personal projects after a while they becomes a giant messy ball of wires and i basically trust the model to untangle it for me , by the time it untangles properly, I run into my token limits.
With Opus 5 / Fable / gpt-5.6 you can simply ask them to fan out subagents to look for ideal architectural simplifications and rank their findings by impact vs confidence.
You can swap out "architectural simplification" with performance opportunities, bugs, correctness, etc. I get the orchestrator agent to then itemize it all into a file where I can keep track of which ones I've implemented.
The results are pretty astounding. I run these right before my weekly limits reset for each subscription and the findings will dictate the secondary tasks I get done during the week.
It's definitely token-heavy. I'm on the $200/mo Claude Code sub and the $100/mo Codex sub.
But it's pretty clear to me that software engineering is more or less solved and all you need is enough patience + tokens to get what you want. I think 20 years of engineering experience more lets me save on tokens rather than unlock things nobody else can build.
An example of the scope of one of my AI-engineered projects is a iterm2/ghostty-like terminal app that implements its own pty session, parsing, rendering. It's almost 2000 commits right now.
That said, I have a specific workflow that isn't just a blind "ok now make it so a screen can be split into panes". I have a plan phase focused on coming up with ideal invariants and such. But I'm not sure anymore how much of that is useful vs just yoloing a solution and then paying technical debt in sweeps, like garbage collection.
Every time I try to use a lower tier model like Terra or Sonnet I regret it, so I just use the best one the first time to keep my sanity (it sort of works). Same with effort, you gotta just max it out (though I stopped using Ultra/Ultracode) and I never use fast because I'd rather work in parallel on more than on one thing faster almost always.
Sure OpenAI has open source tools, and equivalent or better models, and lower prices, and they get along better with other agentic tools, but Anthropic has much better marketing and I guess that's what matters.
I don't know why this comment's getting down voted. It's absolutely true. I would also add that Anthropic seems to limit, compute, the more you use it because I'm on the Max subscription, and it just stops working for me.
I'll ask it to do something and it'll say, I tried, but I couldn't do it over and over again or some variation of.
But it doesn't do that at the start of my subscription, so...
Because "$X is bad" is boring, especially when it comes to LLMs. I use Opus 5 and it's fine. It's more chatty and sometimes a little more difficult to steer than 4.x was, but I've found the code quality of 5 to be much higher.
Everyone has different experiences with these things (for example, I've never experienced what you describe), and "$X is bad" is not conducive to thoughtful discussion.
The issue isn't whether it's fine, it's that both Fable 5 and Opus 5 are worse to use for the job of programming than Codex-5.6-Sol in my experience. I am directly comparing these top tier models every week in my work, and many friends have left Claude in favor of Codex recently because of Claude's poor model quality.
One thing I do like about Claude is that the normal (non-Code) chat interface supports MCP, whereas ChatGPT basically does not.
Huh? Comparing model quality is something everyone is constantly doing here. If you need more specifics, it speaks in an incomprehensible way, stops working lazily, does a bad job and misses things constantly, and makes a lot more mistakes. I have both a $200/mo Codex and Claude subscription and swap between them when one runs out, and I'm always more miserable when I have to use Claude.
The Claude desktop app is also widely panned, as I mentioned, and for me this mainly is due to general UX and a poor remote control interface. Codex's connected machine support is top notch.
I have no idea why anyone would spend the money Anthropic asks for Claude models these days. There was a brief window in time when it was worth the money to use them a bit through GitHub Copilot, but there's no way in hell I'm spending that for Claude Sonnet when DeepSeek is basically doing all of my work right now.
Quite a number of the smartest people I know got scalped to work at OpenAI. I don’t know anyone who went to Anthropic. Something I think which is under emphasized is that OpenAI has the human capital in addition to the financial capital advantage over Anthropic.
Over the long term I think OpenAI will produce the better experience when it comes to model quality, harness quality, and availability. I have been using codex the past few months and never looked back.
I think the difference between the Anthropic token maximization approach (vibe code all the things!) and OpenAI's focus on efficiency, terseness and token reduction are going to be the defining features of who wins the long-term race.
My money is on the more efficient solution. Even if Anthropic can win some benchmarks by using 3x tokens over 3x time, it is a terrible base to build toward the future. Users are no longer willing to wait exponentially long for linear improvements. And as we see from some of the Chinese models, they can quickly distill frontier models with the tax of being slower and more token-guzzling, while retaining most of the quality. The real differentiators are becoming speed and efficiency, which translate to cost and user velocity more than incremental capability improvements.
It's great that frontier models can solve complex math equations, but bread and butter LLM usage (where the money is made) has already shifted from "I need the best always" to "what solves my day-to-day problems quickly and consistently". Fable usage as a percentage is flat-lining. We are already at the point where output quality is negligible. What wins going forward is cost, speed, consistency and the compounding effects of "softer" improvements to the harness.
You've articulated what I found unsettling about Boris' pov that "coding is a solved problem". I listened to a few of his talks and was instinctively off-put by that sentiment. I figured a fellow programmer would understand and speak on the nuances.
Granted it did make me think about my biases and to lean into more future facing inevitabilities. But you've nailed it, for Boris and Anthropic, they are betting that coding is a solved problem in the sense that any person can one-shot any random idea and the output will be in some abstract sense "good". And then at what cost and toward what end?
I really would like to see OpenAI’s focus on efficiency but everytime I use Codex, it wastes tokens like there’s no tomorrow, hitting week limit in a day, where I’m able to use Claude just fine. Maybe it’s based on the codebase, I don’t know, but I have better results with Claude than Codex.
same boat.
gpt 5.5/5.6 goes further on its own much more often than opus 4.8/5 does. codex capped ~300k context when claude does 1m.
I don't feel codex is saving tokens, and result is usually not as good imo.
> who wins the long-term race
Isn’t the race between Chinese open-weight models and the others more decisive for the future?
it'll all be centralized models justifying their capability against local inference in, maybe 3 years or less
once we have a bit more memory fab capacity and the insane bottomless investment in AI giants realizes there is a bottom, local hardware will catch up with model performance to the extent that centralized inference will be downgraded to special cases or for orgs that find it cheaper than buying expensive hardware
but really most power users are going to have 1 TB unified RAM and local models that will do well enough
for light office use you can still have a cheap laptop and a claude subscription
Oh no, I use the top shelf models every day and I really think they have a lot of room to improve in pretty much every regard. I suspect Fable usage is flatlining because it's not that good comparitively and way too expensive.
For anyone on the fence, I was a hardcore CC user since it was released. I made the full switch to 5.6 sol and Codex about a month ago.
It's the better experience. The limits are way higher (I almost never burn through my $200/m plan), and the output is better than Opus 4.8 (Opus 5 is completely unusable for me).
Looking like this will be the last month with Anthropic. Between the outages and just overall crap utility of Opus/Fable lately...
They'll all be getting around to this sooner or later, its just too expensive
I get incredible utility out of $200 per month
It's hard to put a number on it, but even accounting for all the time in meetings, talking with stakeholders and developers etc I'm over 10% more productive overall. I earn substantially more than $2000/month, the ROI is there
It's only expensive compared to the currently very strong offering from OpenAI. Or other models - my hobby projects are all on DeepSeek
I posted on LinkedIn that I thought the future of AI will be local. I'm surprised I didn't get tarred and feathered, but it's been crickets.
I do believe it though. While the subsidies of the frontier providers are nice, they can't continue forever.
also codex + sol is really good and quite fast.
I switched when 4.7 dropped, after over a year on the service. It was an immediate improvement - not even 1 hour before Codex was giving me better results with less effort.
Unfortunately Sol does not compare to Fable at all.
And Fable doesn't compare to Sol at all. Fable will block on many request in name of cybersecurity, sol never does that.
Stick with xhigh rather than turbo for most tasks though, now that the Age of Abundant Resets seems to be well and truly over for good (and 5h limits are supposedly returning soon).
Maybe with lower weekly quotas the load will be lower and we'll have fewer outages? A guy can dream.
The promotion is ending; the limits are reverting to pre-promotion levels.
"From May 13, 2026 through August 19, 2026, your weekly usage limit in Claude Code is 50% higher."
During this period however they’ve released models which lots of people report to be significantly more chatty. Framing it purely as a promotion ending feels like you’re giving them a bit too much credit. I doubt it’s a coincidence they started this promotion a week before the release of 4.8.
I think running a model unsupervised in anything above medium is a sucker move to burn more tokens. The high effort models can be great in limited context, but unsupervised they too often end up navel gazing. High doesn't always mean smarter, but it always burns more tokens. Medium and low seem to be decent for day to day tasks.
I am not sure the supervision is the problem, it's more the way Claude is acting in general, taking initiative that could burn thousands of token in a second. Just today I ask it a small fix on a newly pulled project, it wrote 5 lines of code but then immediately started to run the full test suites, which itself runs in docker, with several images that I didn't pull yet so it started to download everything and by the time I click stop, it already filled the context windows with dozen of hundreds of useless lines... I find it very hard to fins the sweet spot in the prompt between "do not assume, do not take initiatives" and "inform me if there is a simpler way to do what I am asking or if there are some downsides I may not have thought about".
That’s my experience too, going off the deep end with CoT doesn’t meaningfully improve output on a lot of tasks but it does incinerate tokens and take longer
My favorite part (codex thing) is how it runs all commands twice, every time.
Once on a sandbox, realizes nothing works in its sandbox, then again outside the sandbox.
Fucking hell.
I found the waffle reduced my overall usage. It doesn't matter much if it's smarter if it also needs more babysitting. It's quite good at making something quite advanced that is tangentially related to what you actually wanted.
As soon as it gets annoying enough to switch to another provider your Claude code tokens drop right off.
The model has a new improved tokenizer that also counts the same outputs as 20% more tokens
Choco rations going up.
Double plus good, eh!
Anthropic seems cooked right now in terms of compute
Flavor of the month LLM gets inundated and swamped with users.
Sideline LLMs have free compute and offer cheap prices to draw in crowd. Becomes flavor of the month LLM.
Back to step one.
It should be pretty clear by now that token prices are predominately a function of available compute.
They've been really struggling all of 2026. All of these "limited time" promos, just to have less usage than codex, and the shenanigans around "peak hour" reduction earlier in the year.
To be fair, their growth is ridiculously insane. I don't know how else that can be described
Meanwhile my $100/mo subscription allows me to spank Claude all day doing the work of 4 coders and never hit any limits.
This is my experience as well -- unless I'm using Fable. That one burns through tokens like a lightsaber through butter.
It’s like we’re hot-spotting, but across providers rather than across infrastructure!
Currently (and regularly) down, indeed.
it looks cooked to you, because growth numbers are cooking (to them)
(nobody uses claude anymore, it's too over-subscribed)
I got a max account during this promotion and it's been very fun but I'm a little burned out and am kind of looking forward to going back and tinkering with game engines without the help of an LLM for code gen (will still use it for documentation questions but I can do that with the free tier)
For fun/side stuff, handwritten code makes total sense, if you’re burnt out.
For anything that makes money, it feels like a huge step down in productivity, even with the downsides of agent-assisted code.
Oh yes I still use it at work.
The max account is for home projects where I'm basically making some common tools but tailored to myself (diet tracker, note app etc).
It's been great but across work and home it's just too much.
Ah, forgot about that. I was questioning if I should switch my big subscription to ChatGPT (was on the 20$ Open AI and 100$ Claude), seems like a no brainer now
I've been watching this to see if they "extend" it again, let it end, or make it permanent. I'm regularly hitting 90-100% on my $200/mo sub and will switch to Codex in a heartbeat if they drop these limits. This constant uncertainty is really annoying.
Anthropic's struggles have only started - https://s-1.vercel.app/posts/the-capital-cycle-theory/
Will they, though? I could see Anthropic extending it since OpenAI cut the pricing of Sol by 50% for the time being.
Sure seems like OAI is using their compute/capital advantage to dump. Seems improbable they're making a profit at those levels.
Why, it seems like the supply is limited by the ability to provide, and the demand is limited by market alienation.
They raise their prices too high, a lot of customers will still buy but be unhappy about it, it's burning goodwill for money. If they drop their prices too low they have overwhelming demand.
They could be making money hand over fist for all we know. We don't really know how much compute is being used, nor how much it cost them.
To anyone like me that switches between Codex and CC based on limits, but prefers CC to Codex, PI has been really great to migrate over. It can customize itself very easily, I was expecting to have to clone the repo or whatever, but no you can just tell it "add a manual mode where you present all changes as diffs in VSCode, letting me edit them or save to accept" and it will do it.
I wish I could use my Anthropic sub with it but I heard you get banned, but at least you can use it with any other subscription or model.
I suspect the net result will be that more Pro $100 subscribers upgrade to $200/mo. If you explicitly do simpler stuff with Opus, you generally have enough Fable time for most programming and planning tasks.
$100 Claude and $100 ChatGPT Pro is the best value for $200/month. They can see each others' mistakes.
Not to mention a lot of people who already have more than one max20 sub getting additional ones.
prediction: extension of promotions because of the progressive move to codex.
With how bad Opus 5 and how expensive and limited Fable is, I would not be surprised.
Are there any figure to back up people moving to Codex?
I’ve been using both at the same time, but fable had me hitting my limits; even less will have me get 2x codex max instead of one of each.
No. Here's a data point: I tried it for a few days. While it worked fine I didn't see any benefit to switching since it costs the same ($100) and I've spent a year getting up to speed on Claude usage and quirks.
I for one, and I know of plenty others.
I wonder if they schedule these promotions around training runs?
If Anthropic doesn't figure out why OpenAI is eating their lunch, they're in some trouble. I'm guessing they already know why, but aren't able to accept that they won't somehow come out #1. Or in fact they really believe they will still come out #1. Delusion?
Was anyone impressed by Fable? I was never wowed by anything it generated.
It's blowing my mind. Fable is the only AI model I use for coding, everything else is garbage for my use case.
This and the word watermarking is rubbing a lot of people the wrong way, me included.
Like, stop toying around with token limits and just focus on more efficient models.
Once the local models are good enough we are so abandoning these elephants.
I'm gonna walk as soon as possible.
I wonder if this promotion was because of the summer holidays.
Does summer holidays only apply to Europeans? I think most Americans are still largely in office for August
People take much more time off during the summer in the USA than during other times of year. It’s somewhat regional, but it’s been true in all the place I have lived. Probably July is the most common month though because in some regions schools start sometime in August. It’s not uncommon for people to take 1-2w off during that period. More than 2w often requires extra approvals, so is less common.
Exactly. Say on average people take 2 weeks / 16 week off in that period, thats 12% traffic right there.
And smarter people than me can probably find some extra-linear relationship between usage and required limits for uptime etc.
I saw this announcement when it came out and completely forgot about it. That explains why Claude Code felt so surprisingly generous these past few months.
I forgot this was in effect. I’ve been hitting my weekly rate limits two to three days in because of needing to steer Slopus 5 with Fable. Absurd that they’re being cut further in the face of steep competition from open weights models.
On the same day that OpenAI cuts per token cost by 50% on GPT5.6 Sol. :')
they never made money on Claude code, it was always a way to get used recommend API access to your company
This might explain the outtages today, as folks race, misguidedly or not.-
Buy a subscription, get something different every month.
I actually use claude these days mostly for mission critical tasks and code reviews. Everything else I use Sol. I have $20/month plans on both. For non-coding I also use Gemini on a $20/plan. So far I haven't had any issues but we'll see how things go. I have Qwen3.8-27B installed locally but until I upgrade my mac it's not for day to day stuff.
As we speak, Claude is down. I was already complaining about how extremely slow and dumb it has gotten since Opus 5. I guess this will be the last nail in the coffin for Claude. At least for me.
i'm using claude right now, like literally in an iterm2 window next to my browser window where I'm typing.
The 'tokenmaxxing' and the extreme gambling fiesta with Anthropic's latest Claude Code slot machine engine called 'Fable 5' which has given many subsidies and free spins of the wheel and cheaper tokens are all incompatible with the desires of their future owner: Wall Street.
If Wall Street sees a single outage or a tiny drop in usage, they won't be happy and will pressure Anthropic to take away the free tokens.
Better to reduce the limits now rather than to wait until Wall St. tells them to just to avoid a stock punishment.
And in the end, selling tokens ENCOURAGES the companies to sell more tokens. And how do they sell more tokens?
"Thinking" aka 'trust us bro!' without proof of thinking.
Making the model waste more tokens.
Advertisement: aka you pay to be advertised at.
Silently downgrading you and still faking models with the more costly tokens.
Intentional strategies to eat more tokens with no real gains.
Giving out almost-but-not-quite solutions that require another pull of the slo(t/p) machine.
Claude Opus and Fable are so bad compared to GPT-5.6-Sol it's ridiculous, their desktop client is worse, and the value is worse because OpenAI has been spamming discounts. They better get their shit together at Anthropic!
Edit: If you're mad I'm light on the details I have provided some in the replies.
gpt-5.6-sol uses fewer tokens and OpenAI has been more generous with usage, but quality-wise I wouldn't say they are inferior to each other.
I have a Claude Code and OpenAI subscription so that I can use Opus/Fable/gpt-5.6 as I please, and the models are often catching things the other models missed. So much that I would significantly weaken my workflow if I dropped one subscription.
My best workflow at the moment is to create the initial plan with Fable (before review/revise-cycling with other models). From my own testing it seems slightly better at arriving at high-level ideal solutions after sweeping the whole project, projecting future needs, then coming up with good trade-offs like "by construction" correctness.
While mostly subjective, maybe the closest objectivity I have here is noticing fewer revision cycles needed with Fable-initialized plans.
Any model currently in 2026 is better than the ones of 2024. What is the quality difference really.
I mean i think if a developer has a good handle of the code the difference is marginal .
Unless we 100% offload the thinking to the model and act like a prompt manager. Maybe
For personal projects, I defer almost everything day-to-day to the models, so it gives me a good feel for what the models are good at.
Even then, it's kind of a wash these days between the sota models, and we're talking about maybe a 10% performance difference or something. But every once in a while there's the experience of one model spinning its wheels on a bug/repro/issue while another model comes in and one-shots the solution.
That make sense. Let me ask you this When the project reaches a level of complexity . Do you simply reach for better models or do you reengineer it or does the project scope stop at the egdge of the model's capabilities.
I am asking because in my personal projects after a while they becomes a giant messy ball of wires and i basically trust the model to untangle it for me , by the time it untangles properly, I run into my token limits.
With Opus 5 / Fable / gpt-5.6 you can simply ask them to fan out subagents to look for ideal architectural simplifications and rank their findings by impact vs confidence.
You can swap out "architectural simplification" with performance opportunities, bugs, correctness, etc. I get the orchestrator agent to then itemize it all into a file where I can keep track of which ones I've implemented.
The results are pretty astounding. I run these right before my weekly limits reset for each subscription and the findings will dictate the secondary tasks I get done during the week.
It's definitely token-heavy. I'm on the $200/mo Claude Code sub and the $100/mo Codex sub.
But it's pretty clear to me that software engineering is more or less solved and all you need is enough patience + tokens to get what you want. I think 20 years of engineering experience more lets me save on tokens rather than unlock things nobody else can build.
An example of the scope of one of my AI-engineered projects is a iterm2/ghostty-like terminal app that implements its own pty session, parsing, rendering. It's almost 2000 commits right now.
That said, I have a specific workflow that isn't just a blind "ok now make it so a screen can be split into panes". I have a plan phase focused on coming up with ideal invariants and such. But I'm not sure anymore how much of that is useful vs just yoloing a solution and then paying technical debt in sweeps, like garbage collection.
Every time I try to use a lower tier model like Terra or Sonnet I regret it, so I just use the best one the first time to keep my sanity (it sort of works). Same with effort, you gotta just max it out (though I stopped using Ultra/Ultracode) and I never use fast because I'd rather work in parallel on more than on one thing faster almost always.
Sure OpenAI has open source tools, and equivalent or better models, and lower prices, and they get along better with other agentic tools, but Anthropic has much better marketing and I guess that's what matters.
I don't know why this comment's getting down voted. It's absolutely true. I would also add that Anthropic seems to limit, compute, the more you use it because I'm on the Max subscription, and it just stops working for me.
I'll ask it to do something and it'll say, I tried, but I couldn't do it over and over again or some variation of.
But it doesn't do that at the start of my subscription, so...
Because "$X is bad" is boring, especially when it comes to LLMs. I use Opus 5 and it's fine. It's more chatty and sometimes a little more difficult to steer than 4.x was, but I've found the code quality of 5 to be much higher.
Everyone has different experiences with these things (for example, I've never experienced what you describe), and "$X is bad" is not conducive to thoughtful discussion.
The issue isn't whether it's fine, it's that both Fable 5 and Opus 5 are worse to use for the job of programming than Codex-5.6-Sol in my experience. I am directly comparing these top tier models every week in my work, and many friends have left Claude in favor of Codex recently because of Claude's poor model quality.
One thing I do like about Claude is that the normal (non-Code) chat interface supports MCP, whereas ChatGPT basically does not.
> I don't know why this comment's getting down voted.
Because just calling something bad does not add a lot to the conversation. It's not thoughtful, interesting, or good.
Huh? Comparing model quality is something everyone is constantly doing here. If you need more specifics, it speaks in an incomprehensible way, stops working lazily, does a bad job and misses things constantly, and makes a lot more mistakes. I have both a $200/mo Codex and Claude subscription and swap between them when one runs out, and I'm always more miserable when I have to use Claude.
The Claude desktop app is also widely panned, as I mentioned, and for me this mainly is due to general UX and a poor remote control interface. Codex's connected machine support is top notch.
I also mentioned the value of the Codex resets!
It might be objectively true for you.
Opus 5 is fine for me and works better and faster on low and medium than higher effort on prior versions. Same as 5.6 Sol compared to 5.5 or 5.4.
I have no idea why anyone would spend the money Anthropic asks for Claude models these days. There was a brief window in time when it was worth the money to use them a bit through GitHub Copilot, but there's no way in hell I'm spending that for Claude Sonnet when DeepSeek is basically doing all of my work right now.
The charge me exactly the same: $100.
Will Anthropic be the Netscape or Yahoo of our time ?
Yet another scramble to use up weekly quota before an Anthropic cliff. This isn't fun.
Quite a number of the smartest people I know got scalped to work at OpenAI. I don’t know anyone who went to Anthropic. Something I think which is under emphasized is that OpenAI has the human capital in addition to the financial capital advantage over Anthropic.
Over the long term I think OpenAI will produce the better experience when it comes to model quality, harness quality, and availability. I have been using codex the past few months and never looked back.