Claude Code Quietly Got 5x More Expensive

I started with an easy quota question. Should I get another Claude account or have I been using the ones I have poorly? Then I counted the logs and the answer was much worse.

I'm writing this while working on vroni.com, my AI coding tool for turning tasks into pull requests.

The price per month remained unchanged. The machine was identical. This was the kind of work I've been doing for weeks. Claude Code, review loops, long running agent jobs, local automation runs. But the amount of Opus work you can see before your account starts to feel empty is nowhere near what it used to be.

So here is the dull figure that is most important to me: visible input and output from Opus. No cache read, only the tokens where Claude read or wrote normal text. 1,442,423 tokens in this week's rolling seven-day window. Same machine, two previous heavy weeks, 8,927,831 and 8,477,395.

So the current week is 16.2% and 17.0% of those old weeks. A sixth maybe. No, that's not "I used it a bit differently."

That is the same subscription looking like it provides dramatically less of the thing I pay for.

If you want to be generous and include cache creation, that's fine. Creating cache is real work. Cache reads are different things so I still leave out cached reads. Current week for that broader non-cached metric was 4,563,761. The old heavy weeks were 22,155,800 and 23,077,312. That means the current week is 20.6% and 19.8% of the old ones.

So the strict count says there's about 83% fewer. The generous count says about 80% less. In either case, the effective price of the useful work seems to be about 5 times higher.

What I can't explain away is the fresh account from June 29, 2026. It was basically full at about 13:44 CEST. 5h:100% and 7d:100% or so. By around 22:45 CEST it showed around 5h:65% and 7d:84%.

In that period there are 0 visible raw Opus rows.

That's not to say a metered event didn't happen somewhere in Anthropic. That shows the user-visible counting didn't explain the meter. That is the problem.

When I did a real benchmark burst later, the logs did reflect that. That run had 1,246,550 input+output, 5,052,266 cache_create, 6,298,816 non_cache_total, and about $108.55 of raw logged cost.

So no, this is not me making out like heavy usage never happened. The logs reflect the heavy usage. The complaint is the meter collapse and the visible accounting doesn't add up.

This also hits hard as Anthropic publicly announced increased Claude usage limits on May 6. Then we had June, full of Claude status noise, like the June 30 Opus 4.8 incident from 14:31 UTC to 15:28 UTC. And the Reddit usage-limits megathread is full of descriptions of sudden drain and mysterious weekly-meter behavior.

I have the receipt, my local logs.

There could be a sensible internal rule. Maybe failed calls count. Maybe retries matter. Maybe Opus 4.8 is differently weighted. Perhaps Claude Code uses up its quota before the local stats file logs a row. But if that's true, Anthropic has to show it.

Developers don't need a fancy progress bar. We need a ledger. Timestamp, model, surface, input, output, cache_create, cache_read, success or failure, retry count, five hour meter delta, seven day meter delta.

That ledger protects Anthropic if the meter is fair. If the meter got harsher, it protects users. From the outside, Claude Code looks like the same monthly subscription was roughly five times more expensive for the work I actually care about.

![Image 6: VroniWhat I'm building ## Delegate tasks. Get software. Give Vroni a GitHub issue, bug report, spec, or rough idea. It reads the repo, plans the change, writes code, runs checks, and works toward a review-ready pull request. Take a look at vroni.com!Image 7](https://www.vroni.com/?utm_source=vincentschmalbach.com&utm_medium=blog&utm_campaign=blog_post_vroni&utm_content=after_post_card)

Subscribe to my newsletter

Get new posts when I publish them.

I respect your privacy. Unsubscribe at any time.

One thought on “Claude Code Quietly Got 5x More Expensive”

1. !Image 8**q**says: June 30, 2026 at 3:35 pm Maybe your machine has been shai hulud’ed? Reply

A blog that gets read

I write about software, AI, Laravel, SEO, and what it takes to build online products. In 2025, 73k people read the blog. Some posts were picked up by publications and developer communities.

Read the blog Browse Laravel posts

73k readers in 2025

241 published posts

919 comments

Featured posts

A few older posts got the most attention. I keep them here, separate from the newest writing below.

[SEO / Google ### Google Now Defaults to Not Indexing Your Content The post that led to a Guardian column and a long developer-community discussion. Read post](https://www.vincentschmalbach.com/google-now-defaults-to-not-indexing-your-content/)

[AI / Engineering ### AI Exponentializes Your Tech Debt Why AI makes weak engineering foundations more expensive, not less. Read post](https://www.vincentschmalbach.com/ai-exponentializes-your-tech-debt/)

[AI / Development ### Only Experts Can Write Good Prompts Good AI work still depends on technical judgment and taste. Read post](https://www.vincentschmalbach.com/only-experts-can-write-good-prompts/)

Laravel ### Do Not Use Laravel Cache Tags Specific Laravel advice for people working in real codebases. Read post

Latest posts

The newest posts from the blog. This updates automatically when I publish.

All blog posts

[Jun 30, 2026 / Claude Code ### Claude Code Quietly Got 5x More Expensive I started with an easy quota question. Should I get another Claude account or have I been using... Read post](https://www.vincentschmalbach.com/claude-code-quietly-looks-5x-more-expensive/)

[Jun 30, 2026 / API ### Claude Code Is Quietly Fingerprinting China-Linked API Routers Claude Code has a hidden test for where your requests are going. The trigger is ANTHROPIC_BASE_URL, the environment... Read post](https://www.vincentschmalbach.com/claude-code-china-router-fingerprint/)

[Jun 27, 2026 / LLM ### Small Teams Will Ship More Software Than They Can Maintain Using Cursor and Claude, a 5-person engineering team can now build a billing webhook handler, a customer import... Read post](https://www.vincentschmalbach.com/small-teams-will-ship-more-software-than-they-can-maintain/)

Browse by topic

Laravel 71 AI & LLM 65 Software Development 16 Git 11 Business 10 SEO 7

Known from

Places that have mentioned, quoted, or discussed my writing and commentary.

!Image 9: The Guardian

!Image 10: Forbes

!Image 11: Business Insider

!Image 12: Hacker News

What My Clients Say

Actual feedback from clients I've worked with.

- * *

> "HIRE VINCENT. Seriously. I've worked with many many developers over the years (remember rentacoder?), and Vincent is that rare breed of web developer who just gets it. You have a chat, you express what you need, and he comes back with what you asked for. His english is fluent, he understands what you need, and he's not looking to overcomplicate things. It's a joy and pleasure to work with him, not to have to micromanage and just overall a great experience. Honestly I wish I could keep him on we just don't have the budget anymore - that is the only reason the contract ended." > > Ariel Ozick Wired Rhino, Inc.

> "What stood out about working with Vincent was how methodical and efficient he was. He started by really digging into our requirements, asking clarifying questions until he had a solid understanding. He then built exactly what was needed and kept us informed throughout. Communication was clear and proactive at all times. Everything stayed on track, both in terms of time and budget. The end result was excellent and has been running smoothly since launch. It's that combination of practical focus and attention to detail that makes the difference. I would definitely recommend Vincent for quality development." > > Philipp Toepelmann Digital Investments GmbH

> "Having Vincent on retainer has been excellent. He's tackled some major technical challenges for us, notably significantly improving our website's performance. He also expertly managed a complex upgrade of our core platform (Laravel) to the latest version, ensuring we stay current and secure. We're now rolling out the multilingual capabilities he built, which is exciting. Vincent is reliable, technically skilled, and a valuable partner for keeping our site running optimally and moving forward. Highly recommend his expertise." > > Howard Rickard Athena Advisers Ltd

> "I've been working with Vincent for several months, and he's one of the most independent developers I've collaborated with. He doesn't need micromanagement - he's fully capable of making decisions and driving the project forward, which has freed me up to focus on other parts of the business. > >

> Vincent works very well asynchronously: communication stays focused, there is little unnecessary back-and-forth, and he still raises the important questions when needed. The result is a developer who delivers without overhead, and that's been genuinely valuable." > > Aidas K. Bizdras OU

> "We work with Vincent regularly on Laravel/Vue projects, both new builds and existing codebases that have grown over the years. In both cases, he delivers clean code and reliable implementation. We especially value how quickly he gets into mature projects and finds solutions that fit the existing state of the codebase. Absolutely recommended." > > Florian Brunner designundzwanzig OG

- * *

> "Vincent was easy to work with and produced high-quality code. Would recommend." > > Dustin Fraker QuickerNotes, LLC

> "Vincent is one of those rare developers who can jump into a complex codebase, become useful immediately, and solve real problems without hand-holding. > >

> He worked on performance optimization, customer-specific features, and a major custom reporting system. He also turned customer discussions into clear technical requirements and documented his work so the rest of the team could build on it. > > > Vincent does not just close tickets, he thinks through the product, the edge cases, and the customer impact. If you need a senior Laravel developer who can move a SaaS product forward, I strongly recommend him." > > Yash Chandra TREFT SYSTEMS INC DBA Academy of Mine

![Image 19: Vincent Schmalbach](https://www.vincentschmalbach.com/about/)

Have something I should look at?

Send the rough version. I can help turn it into a plan and then build it.

Get in Touch Explore my services

![Image 20: Rapid SaaS with Laravel Book Cover](https://www.rapid-saas-with-laravel.com/)

My Book: Rapid SaaS with Laravel

Launch Your SaaS in Days, Not Months. A practical, no-BS guide to building production-ready SaaS apps with Laravel 12: speed, billing, teams, AI integration, and the parts you need before customers can pay you.

Visit Book Homepage Get it on Leanpub

#### About

Vincent Schmalbach is an entrepreneur, software developer, and SEO expert. He works on SaaS products, business apps, automation, and technical project rescue, with 15+ years of software and online business experience.

[](https://www.linkedin.com/in/vschmalbach)[](https://twitter.com/vschmalbach)[](https://www.xing.com/profile/Vincent_Schmalbach)[](https://github.com/vincentsch)[](https://angel.co/vincent-schmalbach)

#### Posts

- Claude Code Quietly Got 5x More Expensive

- Claude Code Is Quietly Fingerprinting China-Linked API Routers

- Small Teams Will Ship More Software Than They Can Maintain

- AI Makes Bad Product Decisions Look Like Finished Software

- The AI Productivity Trap Is More Output

- If Software Gets 90 Percent Cheaper, Who Captures the Savings?

- The AI Coding Era Makes Boring Tests More Valuable

- Open Source Maintainers Need a Spam Filter for AI Labor

- The Junior Developer Problem Is Becoming a Senior Developer Problem

- Vibe Code Is Legacy Code Before It Ships

- AI Code That Passes Tests Can Still Fail Review