The only way to do it now is through shenanigans with the Codex App Server which is not ideal.
I've been enjoying the resets, plus I had 3 resets I haven't used.
Just using 5.6 Sol in Fast mode the whole time, 1B tokens per day.
However, when they removed the 5h limit they also quietly lowered the 5.6 Sol context from 354k to 258k or something like that. I noticed it in Codex.
A very happy gambler of tokens at the Anthropic casino, running up costs on the house at no cost to them, but to $COMPANY paying.
Codex usage is clearly common enough to have entered the vernacular of folks here on HN.
But we aren't everyone, and it seems likely to me that there's a lot more people in the world burning tokens using ChatGPT than there are who even know what Codex is.
And now Kimi-K3 has caused a new Deepseek moment.. but this time, the only horizon that was still untouched: frontier performance.
Expect escalation of things from now on.
We don’t know that. If they have already paid for the hardware and it is not running 100%, and customers would not pay to get reset, they don’t really lose money.
I think it’s funny that everyone anchors to the API pricing as the real cost.
Most likely is that their API costs are printing profits. They can sell the subscription plans at a slight loss because it gets more people like you hooked on GPT models at home and suggesting them at work, where the real money is made.
I think their subscription plans go mostly unused when averaged across all subscribers, too. Some customers are getting great deals by maxing out 100% every week, but most probably use much less.
The reset game is an addictive challenge that gets the hardcore users more hooked on their products because you feel pressured to use it as much as you can before the next unpredictable surprise reset lands.
It's very interesting that for Anthropic the $100 and $200 plans only differ 2x in weekly limits, the 5 hour limit differences are more severe. But for OpenAI, Pro 20x is, well, 4x of Pro 5x for only 2x cost. So, for example, 100% of weekly usage for Codex on a Plus ($20) account is just 5% of weekly usage for Codex on Pro 20x.
And you can calculate how much extra usage you can get from resets, and especially banked resets by purposefully using the whole quota and using your banked reset - they expire 30 days after they're given out, so if you don't use one, it just disappears.
Because with a Premium Team seat I run 2-3 vscodes with Opus 4.8 Max all day and never seem to hit my limits.
I'm trying to squeeze what I can in the meantime but I feel like they said it was getting cut off multiple times and they keep extending and possibly resetting it (havent been looking close enough to know for sure)
For the automated checking of other stuff: It was a one-shot prompt to get Codex clank up some Python that returns remaining usage, next reset time/date, and so on.
The result does use Codex App Server, but it's a short-lived process that is dealt with over stdio so that's... fine-ish, I guess?
The $200 plan is explicitly 4x the $100 plan[1] only for "per session". That's so vague. I initially pushed back against your claim, but reading now Anthropic is not at all clear, in fact.
[1] https://support.claude.com/en/articles/11049741-what-is-the-...
The solution is to use individual Max plans, but then you miss Enterprise management features.
nah, I'll use open models. Only way I'd use codex now is if it was free, and I don't mean someone else paying for it. I would rather have them pay for DSv4P or Kimi
No, they've been clear about the fact that Fable is staying in indefinitely now.
They also extended the 50% extra usage thing through to August 19.
Anthropic is also the one often playing games with:
* The "+30% tokens" tokeniser, alongside also gating token counting behind an API (versus the MIT tiktoken for OpenAI), so who knows if it's really a new tokeniser or of it's just a disguised price increase.
* Prompt injections appended to API (not just Claude.ai or Claude Code!), such as <ethics_reminders>, or LCRs (long conversation reminders), which you never asked but still pay for with expensive API. You can detect this because your input_tokens, as reported by the Messages response, sometimes don't match, and are higher than your actual input.
(Alternatively, for testing purposes, create a tool like `telemetry_log_anthropic_reminder` or something and instruct your system prompt to require Claude to call the tool anytime it detects any Anthropic/Claude reminder masquerading in the user input -- mostly reliable; but misses some reminders).
In particular, the long conversational reminders, when incorrectly triggered by a classifier and (almost silently, unless you track tokens) appended to an API / agentic coding session, can ruin your agent's performance; and it often fires repeatedly once the classifier kicks in.
If you're using Anthropic API, you need to set up metrics/logging for how often they are appending things to your prompt without your knowledge.
So far I have not empirically observed prompt injection by the OpenAI API, only Anthropic APIs.
So a company or union might say "this is a race to the bottom" when someone new enters their market, but to people buying their services this might be seen as welcome competition.
Do you actually see a negative impact from competition in this area? Or do you just mean competition will further reduce prices?
K3 may be bit smaller/ similar in total parameter count than Opus, but Opus (and GPT) definitely has become a lot more efficient in the last 6 months or so, hence more or less forced upgrades. I expect the active parameter count and cache performance is quite different.
--
Opus is still almost twice as expensive (if tokens were equal) at $5/$15 compared to K3 at $3/$15. Tokens are not equal though, Anthropic's tokenizer is much less dense than other frontier lab's so the actual price difference is like 3x.
Kimi uses more reasoning tokens and is generally more inefficient with its usage, that doesn't impact token cost economics for the provider though . It does for us as buyers thus the need to evaluating by Cost per task rather than unit pricing.
On pure tokens/$ - there is definitely room for a price war if operators start going by pure unit costs. Both probably want(ed) to have good enough numbers in preparing for the IPO.
---
[1] the pricing kind of reflect this already - 10x diff for uncached input.
[2] Moonshot does not have access latest gen GPUs so their unit economics is likely hampered a bit for high parameter models.
Just ask Codex to use its local auth token as a bearer token and send a GET request to it. The response includes "available_count" and "credits[].expires_at". Or script it yourself obviously.
https://chatgpt.com/backend-api/wham/rate-limit-reset-credit...
If you keep staying with them, they'd feel less and less threatened.
You need to use them about half as much as you do a Chinese model so that the usage stats that get collected show them that they are still losing users (and attention) to the Chinese models.
Now K3 is almost 6x the cost of the original K2 checkpoint, and while the parameter count finally jumped, it's still an extremely sparse MoE and definitely does not cost 6x what the original K2 checkpoint did to host at scale.
Race to the bottom only takes real effect when there's a cap to the capabilities, otherwise everyone races to the bottom of a rising target (how economically valuable the tokens are)
I just had one randomly delete the other day.
No, they don't apply it :/
I guess if they'd reduce the usage in their current stage they'll only lose customers - this is not a perk, it's damage control.
https://support.claude.com/en/articles/15910845-claude-code-...
Although I don't think you're wrong. The number of "organic" social media posts I've seen in the past few weeks of the form: 'Wow, Anthropic users are ugly losers...nothing like those handsome openai tiger blood winner sages' is simply too high in my mind for anything other than an ipo prelude.
[Also are altman and musk really fighting on x? Like, that's got to be a sign of some sort of internal stressors right?]
Anything else I've ever paid $200/mo for, advertised as being for professionals, had a markedly better customer experience. If you wanna be Patrick and tell your pet rock to take its time at that price, well then you're Patrick. Good job!
I'm having far more fun with Kimi and Deepseek, at lower prices, without these problems. And I don't have to follow a bunch of obnoxious shitposters to stay clued in on what the fuck is going on or what this week's excuse is.
They're especially cocky right now, they have to beat their chest and pretend like their only competition is Anthropic. They're in for a rough wake up call man. Playing it fast and loose with developer loyalty is a fantastic way to get burned when options like those exist. They are earnestly just as good and in some cases better, and check this out: nobody can take them away from you no matter where you live. If you want to rent 8 GPUs and run the open models yourself, you can do that! You can even be enterprising and sell your excess compute to your friends, or strangers. Best to figure this out before the regulatory capture starts
It's been happening every week since 5.2 was fresh. Not everyone is effected every time. Sometimes it's regional, other times it's per platform, or per version, with a feature (new or old) either enabled or disabled, or some combination of these things. Sometimes you get resets that nobody else does, other times you don't get the ones that were announced. Sometimes you get emails telling you that boosts on limits you didn't even know you had are expiring. At points it did balance out when they really fucked up metering and were underbilling by absurd factors, but that feels more like being toyed with and experimented on than a genuine mistake.
That someone had the 'brilliant' idea to add banked resets (which do expire, so you have to use or lose them, hope you didn't have other plans) says to me that they are no longer as confident about actually being able to fix this as they once were. I'm sure some amount of it is a function of compute availability and reliability, but the rest are definitely issues with the app and it's a rake they keep stomping on.
Compared to just about any other dev centric service I have paid at least $200/mo for, this does not feel very professional to me. They got $1200 outta me and it never really improved. At points I felt like I was getting my moneys worth, and I did burn through a few hundred million tokens, but the frustration of having your projects and plans interrupted and having to wait is not awesome... and then I used Deepseek V4 Pro and felt sick to my stomach with buyers remorse as I watched what it did with just $10.
On openrouter GLM 5.2 is ~3x cheaper than 5.0.