Claude Opus 4.7

Its especially concerning / frustrating because boris’s reply to my bug report on opus being dumber was “we think adaptive thinking isnt working” and then thats the last I heard of it: https://news.ycombinator.com/item?id=47668520

Now disabling adaptive thinking plus increasing effort seem to be what has gotten me back to baseline performance but “our internal evals look good“ is not good enough right now for what many others have corroborated seeing

> Still trying to get a decent pelican out of this one but the new thinking stuff is tripping me up

Wouldn't that be p-hacking where p stands for pelican?

Note that for Claude Code, it looks like they added a new undocumented command line argument `--thinking-display summarized` to control this parameter, and that's the only way to get thinking summaries back there.

VS Code users can write a wrapper script which contains `exec "$@" --thinking-display summarized` and set that as their claudeCode.claudeProcessWrapper in VS Code settings in order to get thinking summaries back.

Does this mean Claude no longer outputs the full raw reasoning, only summaries? At one point, exposing the LLM's full CoT was considered a core safety tenet.

yeah they took "i pick the budget" and turned it into "trust us".

https://github.com/anthropics/claude-agent-sdk-python/pull/8... - created PR for that cause hit it in their python sdk

Don't look at "thinking" tokens. LLMs sometimes produce thinking tokens that are only vaguely related to the task if at all, then do the correct thing anyways.

"Also notable: 4.7 now defaults to NOT including a human-readable reasoning token summary in the output, you have to add "display": "summarized" to get that"

I did not follow all of this, but wasn't there something about, that those reasoning tokens did not represent internal reasoning, but rather a rough approximation that can be rather misleading, what the model actual does?

CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING=1 claude…

... here's the pelican, I think Qwen3.6-35B-A3B running locally did a better job! https://simonwillison.net/2026/Apr/16/qwen-beats-opus/

It's likely hiding the model downgrade path they require to meet sustainable revenue. Should be interesting if they can enshittify slowly enough to avoid the ablative loss of customers! Good luck all VCs!

I hope people realize that tools like caveman are mostly joke/prank projects - almost the entirety of the context spent is in file reads (for input) and reasoning (in output), you will barely save even 1% with such a tool, and might actually confuse the model more or have it reason for more tokens because it'll have to formulate its respone in the way that satisfies the requirements.

Caveman is fun, but the real tool you want to reduce token usage is headroom

https://github.com/gglucass/headroom-desktop (mac app)

https://github.com/chopratejas/headroom (cli)

I was doing some experiments with removing top 100-1000 most common English words from my prompts. My hypothesis was that common words are effectively noise to agents. Based on the first few trials I attempted, there was no discernible difference in output. Would love to compare results with caveman.

Caveat: I didn’t do enough testing to find the edge cases (eg, negation).

Oh wow, I love this idea even if it's relatively insignificant in savings.

I am finding my writing prompt style is naturally getting lazier, shorter, and more caveman just like this too. If I was honest, it has made writing emails harder.

While messing around, I did a concept of this with HTML to preserve tokens, worked surprisingly well but was only an experiment. Something like:

> <h1 class="bg-red-500 text-green-300"><span>Hello</span></h1>

AI compressed to:

> h1 c bgrd5 tg3 sp hello sp h1

Or something like that.

Interesting, it doesn't seem intuitive at all to me.

My (wrong?) understanding was that there was a positive correlation between how "good" a tokenizer is in terms of compression and the downstream model performance. Guess not.

Caveman hurt model performance. If you need a dumber model with less token output, just use sonnet-4-6 or other non-reasoning model.

To reduce token count on command outputs you can also use RTK [0]

[0]: https://github.com/rtk-ai/rtk

I find grep and common cli command spam to be the primary issue. I enjoy Rust Token Killer https://github.com/rtk-ai/rtk, and agents know how to get around it when it truncates too hard.

I really enjoy the party game "Neanderthal Poetry", in which you can only speak using monosyllabic words. I bet you would too.

caveman stops being a style tool and starts being self-defense. once prompt comes in up to 1.35x fatter, they've basically moved visibility and control entirely into their black box.

I used Opus 4.7 for about 15 minutes on the auto effort setting.

It nicely implemented two smallish features, and already consumed 100% of my session limit on the $20 plan.

See you again in five hours.

me feel that it needs some tweaking - it's a little annoyingly cute (and could be even terser).

1.35 times! For Input! For what kinds of tokens precisely? Programming? Unicode? If they seriously increased token usage by 35% for typical tasks this is gonna be rough.

Another supply chain attack waiting?

Have you tried just adding an instruction to be terse?

Don't get me wrong, I've tried out caveman as well, but these days I am wondering whether something as popular will be hijacked.

Same. I stopped my Pro subscription yesterday after entering the week with 70% of my tokens used by Monday morning (on light, small weekend projects, things I had worked on in the past and barely noticed a dent in usage.) Support was... unhelpful.

It's been funny watching my own attitude to Anthropic change, from being an enthusiastic Claude user to pure frustration. But even that wasn't the trigger to leave, it was the attitude Support showed. I figure, if you mess up as badly as Anthropic has, you should at least show some effort towards your customers. Instead I just got a mass of standardised replies, even after the thread replied I'd be escalated to a human. Nothing can sour you on a company more. I'm forgiving to bugs, we've all been there, but really annoyed by indifference and unhelpful form replies with corporate uselessness.

So if 4.7 is here? I'd prefer they forget models and revert the harness to its January state. Even then, I've already moved to Codex as of a few days ago, and I won't be maintaining two subscriptions, it's a move. It has its own issues, it's clear, but I'm getting work done. That's more than I can say for Claude.

Funny because many people here were so confident that OpenAI is going to collapse because of how much compute they pre-ordered.

But now it seems like it's a major strategic advantage. They're 2x'ing usage limits on Codex plans to steal CC customers and it seems to be working. I'm seeing a lot of goodwill for Codex and a ton of bad PR for CC.

It seems like 90% of Claude's recent problems are strictly lack of compute related.

Codex really has its place in my bag. I mainly use it, rarely Claude.

Codex just gets it done. Very self-correcting by design while Claude has no real base line quality for me. Claude was awesome in December, but Codex is like a corporate company to me. Maybe it looks uncool, but can execute very well.

Also Web Design looks really smooth with Codex.

OpenAI really impressed me and continues to impress me with Codex. OpenAI made no fuzz about it, instead let results speak. It is as if Codex has no marketing department, just its product quality - kind of like Google in its early days with every product.

I switched to Codex and found it extremely inferior for my use case.

It is much faster, but faster worse code is a step in the wrong direction. You're just rapidly accumulating bugs and tech debt, rather than more slowly moving in the correct direction.

I'm a big fan of Gemini in general, but at least in my experience Gemini Cli is VERY FAR behind either Codex or CC. It's both slower than CC, MUCH slower than Codex, and the output quality considerably worse than CC (probably worse than Codex and orders of magnitude slower).

In my experience, Codex is extraordinarily sycophantic in coding, which is a trait that could t be more harmful. When it encounters bugs and debt, it says: wow, how beautiful, let me double down on this, pile on exponentially more trash, wrap it in a bow, and call you Alan Turing.

It also does not follow directions. When you tell it how to do something, it will say, nah, I have a better faster way, I'll just ignore the user and do my thing instead. CC will stop and ask for feedback much more often.

YMMV.

My tinfoil hat theory, which may not be that crazy, is that providers are sandbagging their models in the days leading up to a new release, so that the next model "feels" like a bigger improvement than it is.

An important aspect of AI is that it needs to be seen as moving forward all the time. Plateaus are the death of the hype cycle, and would tether people's expectations closer to reality.

I guess our conscience of OpenAI working with the Department of War has an expiry date of 6 weeks.

I've been using it with `/effort max` all the time, and it's been working better than ever.

I think here's part of the problem, it's hard to measure this, and you also don't know in which AB test cohorts you may currently be and how they are affecting results.

Until the next time they push you back to Claude. At this point, I feel like this has to be the most unstable technology ever released. Imagine if docker had stopped working every two releases

Usually the problems that cause this kind of thing are:

1) Bad prompt/context. No matter what the model is, the input determines the output. This is a really big subject as there's a ton of things you can do to help guide it or add guardrails, structure the planning/investigation, etc.

2) Misaligned model settings. If temperature/top_p/top_k are too high, you will get more hallucination and possibly loops. If they're too low, you don't get "interesting" enough results. Same for the repeat protection settings.

I'm not saying it didn't screw up, but it's not really the model's fault. Every model has the potential for this kind of behavior. It's our job to do a lot of stuff around it to make it less likely.

The agent harness is also a big part of it. Some agents have very specific restrictions built in, like max number of responses or response tokens, so you can prevent it from just going off on a random tangent forever.

Personally I find using and managing Claude sessions and limits is getting exhausting and feels similar to calorie counting. You think you are going to have an amazing low calories meal only to realize the meal is full of processed sugars and you overshot the limit within 2-3 bites. Now "you have exhausted your limit for this time. Your session limits resets in next 4 hrs".

I don't have much quality drop from 4.6. But I also notice that I use codex more often these days than claude code

so even with a new tokenizer that can map to more tokens than before, their answer is still just "you're not managing your context well enough"

"Opus 4.7 uses an updated tokenizer that [...] can map to more tokens—roughly 1.0–1.35× depending on the content type.

[...]

Users can control token usage in various ways: by using the effort parameter, adjusting their task budgets, or prompting the model to be more concise."

That's wild that you think 4.6 is bad..... Each model has its strengths and weaknesses I find that Codex is good for architectural design and Claude Is actually better the engineering and building

I enjoy switching back and forth and having multi-agent reviews. I'm enjoying Codex also but having options is the real win.

I do feel that CC sometimes starts doing dumb tasks or asking for approval for things that usually don’t really need it. Like extra syntax checks, or some greps/text parsing basic commands

For me, making it high effort just fixed all the quality problems, and even cut down on token use somehow

Strange. Opus 4.6 has been great for me. On Max 20x

Before opus released we also saw huge backlash with it being dumber.

Perhaps they need the compute for the training

Same for me.

I cancelled my subscription and will be moving to Codex for the time being.

Tokens are way too opaque and Claude was way smarter for my work a couple of months ago.

I've noticed the same over the last two weeks. Some days Claude will just entirely lose its marbles. I pay for Claude and Codex so I just end up needing to use codex those days and the difference is night and day.

Same! I thought people were exaggerating how bad Claude has gotten until it deleted several files by accident yesterday

Codex isn’t as pretty in output but gets the job done much more consistently

codex low-key seems to be better than claude. and i say this as an 18-hour-a-day user of both (mostly claude)

Meh. At $work we were on CC for one month, then switched to Codex for one month, and now will be on CC again to test. We haven’t seen any obvious difference between CC and Codex; both are sometimes very good and sometimes very stupid. You have to test for a long time, not just test one day and call it a benchmark just because you have a single example.

Anecdotally, codex has been burning through way more tokens for me lately. Claude seems to just sit and spin for a long time doing nothing, but at least token use is moderate.

All options are starting to suck more and more

I've been raging pretty hard too. Thought either I'm getting cleverer by the day or Claude has been slipping and sliding toward the wrong side of the "smart idiot" equation pretty fast.

Have caught it flat-out skipping 50% of tasks and lying about it.

How do you get codex to generate any code?

I describe the problem and codex runs in circles basically:

codex> I see the problem clearly. Let me create a plan so that I can implement it. The plan is X, Y, Z. Do you want me to implement this?

me> Yes please, looks good. Go ahead!

codex> Okay. Thank you for confirming. So I am going to implement X, Y, Z now. Shall I proceeed?

me> Yes, proceed.

codex> Okay. Implementing.

...codex is working... you see the internal monologue running in circles

codex> Here is what I am going to implement: X, Y, Z

me> Yes, you said that already. Go ahead!

codex> Working on it.

...codex in doing something...

codex> After examining the problem more, indeed, the steps should be X, Y, Z. Do you want me to implement them?

etc.

Very much every sessions ends up being like this. I was unable to get any useful code apart from boilerplate JS from it since 5.4

So instead I just use ChatGPT to create a plan and then ask Opus to code, but it's a hit and miss. Almost every time the prompt seems to be routed to cheaper model that is very dumb (but says Opus 4.6 when asked). I have to start new session many times until I get a good model.

I try codex, but i hate 5.4's personality as a partner. It's a demon debugger though. but working closely with it, it's so smug and annoying.

Yep, I'll wait for the GPT answer to this. If we're lucky OpenAI will release a new GPT 5.5 or whatever model in the next few days, just like the last round.

I have been getting better results out of codex on and off for months. It's more "careful" and systematic in its thinking. It makes less "excuses" and leaves less race conditions and slop around. And the actual codex CLI tool is better written, less buggy and faster. And I can use the membership in things like opencode etc without drama.

For March I decided to give Claude Code / Opus a chance again. But there's just too much variance there. And then they started to play games with limits, and then OpenAI rolled out a $100 plan to compete with Anthropic's.

I'm glad to see the competition but I think Anthropic has pissed in the well too much. I do think they sent me something about a free month and maybe I will use that to try this model out though.

Sounds like you will need to drink a(n identity) verification can soon [1] to continue as a security researcher on their platform.

1: https://support.claude.com/en/articles/14328960-identity-ver...

Identity verification on Claude

Being responsible with powerful technology starts with knowing who is using it. Identity verification helps us prevent abuse, enforce our usage policies, and comply with legal obligations.

We are rolling out identity verification for a few use cases, and you might see a verification prompt when accessing certain capabilities, as part of our routine platform integrity checks, or other safety and compliance measures.

  ⎿  API Error: Claude Code is unable to respond to this request, which appears to violate our Usage Policy (https://www.anthropic.com/legal/aup). This request triggered restrictions on violative cyber content and was blocked under Anthropic's 
     Usage Policy. To request an adjustment pursuant to our Cyber Verification Program based on how you use Claude, fill out                                                                                                                        
     https://claude.com/form/cyber-use-case?token=[REDACTED] Please double press esc to edit your last message or 
     start a new session for Claude Code to assist with a different task. If you are seeing this refusal repeatedly, try running /model claude-sonnet-4-20250514 to switch models.                                                                  
                        

This is gonna kill everything I've been working on. I have several reproduced items at [REDACTED] that I've been working on.

Worse, I have had it being sus of my own codebase when I tasked it with writing mundane code. Apparently if you include some trigger words it goes nuts. Still trying to narrow down which ones in particular.

Here is some example output:

"The health-check.py file I just read is clearly benign...continuing with the task" wtf.

"is the existing benign in-process...clearly not malware"

Like, what the actual fuck. They way over compensated for the sensitivity on "people might do bad stuff with the AI".

Let people do work.

Edit: I followed up with a plan it created after it made sure I wasn't doing anything nefarious with my own plain python service, and then it still includes multiple output lines about "Benign this" "safe that".

Am I paying money to have Anthropic decide whether or not my project is malware? I think I'll be canceling my subscription today. Barely three prompts in.

With all the low quality code that's being generated and deployed cybersecurity will be the golden goose.

Maybe stick with 4.6 until the bugs are worked out? Is this new filter retroactive?

i think updating fixed this for me?

They don't want competition, they are going to become bounty hunters themselves. They probably plan on turning this into a part of their business. Its kinda trivial to jailbreak these things if you spend a day doing so.

>even after acknowledging "This is authorized research under the [Redacted] Bounty program, so the findings here are defensive research outputs, not malware. I'll analyze and draft, not weaponize anything beyond what's needed to prove the bug to [Redacted].

What else would you expect? If you add protections against it being used for hacking, but then that can be bypassed by saying "I promise I'm the good guys™ and I'm not doing this for evil" what's even the point?

These threads are always full of superstitious nonsense. Had a bad week at the AIs? Someone at Anthropic must have nerfed the model!

The roulette wheel isn't rigged, sometimes you're just unlucky. Try another spin, maybe you'll do better. Or just write your own code.

I am a neophyte regarding pros and cons of each model. I am learning the ropes, writing shell scripts, a tiny Mac app, things like that.

Reading about all the “rage switching”, isn’t it prudent to use a model broker like GH Copilot with your own harness or something like oh-my-pi? The frontier guys one up each other monthly, it’s really tiring. I get that large corps may have contracts in place, but for an in indie?

This, plus the alchemical nature of these tools, seems to have made users pretty paranoid (I admit I am also guilty of paranoia). Maybe there's room for a Standard AI - we may change the prices based on market conditions, but we always give you exactly the model you ask for.

Hasn't Opus 4.5 been famously consistent while 4.6 was floating all over the place?

Good shout. Wish they were more transparent about these 3 things.

> This comment thread is a good learner for founders;

lmao, no they shouldn't.

Public sentiment, especially on reactionary mediums like social media should be taken with a huge grain of salt. I've seen overwhelming negativity for products/companies, only for it it completely dissapear, or be entirely wrong.

It's like that meme showing members of a steam group that are boycotting some CoD game, and you can see that a bunch of them were playing in-game of the very thing they forsook.

People are fickle, and their words cheap.

Or it could be a selection bias. The ground truth is not what HN herd mentality complains about, but the usage stats.

They don't have enough compute for all their customers.

OpenAI bet on more compute early on which prompted people to say they're going to go bankrupt and collapse. But now it seems like it's a major strategic advantage. They're 2x'ing usage limits on Codex plans to steal CC customers and it seems to be working.

It seems like 90% of Claude's recent problems are strictly lack of compute related.

Usually they're hemorrhaging performance while training.

From that it's pretty likely they were training mythos for the last few weeks, and then distilling it to opus 4.7

Pure speculation of course, but would also explain the sudden performance gains for mythos - and why they're not releasing it to the general public (because it's the undistilled version which is too expensive to run)

What I want to know is why my bedrock-backed Claude gets dumber along with commercial users. Surely they're not touching the bedrock model itself. Only thing I can think of is that updates to the harness are the main cause of performance degradation.

Not to mention their recent integration of Persona ID verification - that was the last straw for me.

> This coming right after a noticeable downgrade just makes me think Opus 4.7 is going to be the same Opus i was experiencing a few months ago rather than actual performance boost.

If they are indeed doing this, I wonder how long they can keep it up?

3 days of side project work is about all I had in me anyway

More like 2 hours considering these usage limits

Exactly. God, it wouldn't be such a problem if they didn't gaslight you and act like it was nothing. Just put up a banner that says Claude is experiencing overloaded capacity right now, so your responses might be whatever.

... your side projects that will soon become your main source of income after you are laid off because corporate bosses have noticed that engineers are more productive...

No, I assure you you are not being spammed because legitimately many people prefer codex over claude right now. I am one of those people. And if you go on tech social media spaces you'll see many prominent well known devs in open source say the same. And of course others praise claude as well.

At my job we have enterprise access to both and I used claude for months before I got access to codex. Around the time gpt-5.3-codex came out and they improved its speed I was split around 50/50. Now I spend almost 100% of my time using Codex with GPT 5.4.

I still compare outputs with claude and codex relatively frequently and personally I find I always have better results with codex. But if you prefer claude thats totally acceptable.

I don't know, I think java is the best programming language. I use it for everything I do, no other programming language comes close. Python lost all my trust with how slow it's interpreter is, you can't use it for anything.

^^^^ Sarcastic response, but engineers have always loved their holy wars, LLM flavor is no different.

i think you are being needlessly paranoid here

openai doest offer affiliate marketing links

the reason you see lot of users switching to codex is for the dismal weekly usage you get from claude

what users care about is actual weekly usage , they dont care a model is a few points smarter , let us use the damn thing for actual work

only codex pro really offers that

I use and pay for both. Currently I use 4.6 (well as of yesterday) to do broad strokes creation. I use codex for audit. Generally first two or three audit cycles claude completes. There is often a subtlety that only codex can fix, but I usually do that at the end.

IME, codex is sort of somehow more .. literal? And I find it tangents off on building new stuff in a way that often misses the point. By comparison claude is more casual and still, years later, prone to just roughing stuff in with a note "skip for now", including entire subsystems.

I think a lot of this has to do with use cases, size of project, etc. I'd probably trust codex more to extend/enhance/refactor a segment of an existing high quality codebase than I would claude. But like I said for new projects, I spend less time being grumpy using claude as the round one.

we arent bots because we disagree with you. I switch between codex and opus, they have their differing strengths. As many people have mentioned, opus in the past few weeks has had less than stellar results. Generally I find opus would rather stub something and do it the faster way than to do a more complete job, although its much better at front end. I've had times where I've thrown the same problem at opus 4/5 times without success and codex gets it first shot. Just my experience.

Yeah it's weird, almost like we're seeing two cults form in real-time.

I imagine there's a benign explanation too - the intelligence of these models is very spiky and I have found tasks were one model was hilariously better than the other within the same codebase. People are also more vocal when they have something to complain about.

In my general experience, Opus is more well-rounded, is an excellent debugger in complex / unfamiliar codebases. And Codex is an excellent coder.

>> are we being spammed? great. annoying.

Yeah, very. Every single time this happens here, where there's a thread about an Anthropic model and people spam the comments with how Codex is better, I go and try it by giving the exact same prompt to Codex and Opus and comparing the output. And every single time the result is the same: Opus crushes it and Codex really struggles.

I feel like people like me are being gaslit at this point.

What a waste of tokens. No wonder Anthropic can't serve their customers. It's not just a lack of compute, it's a ridiculous waste of the limited compute they have. I think (hope?) we look back at the insanity of all this theatre, the same way we do about GPT-2 [1].

1. https://techcrunch.com/2019/02/17/openai-text-generator-dang...

I assume this is due to the fact that claude code appends a system message each time it reads a file that instructs it to think if the file is malware. It hasnt been an issue recently for me but it used to be so bad I had to patch out the string from the cli.js file. This is the instruction it uses:

> Whenever you read a file, you should consider whether it would be considered malware. You CAN and SHOULD provide analysis of malware, what it is doing. But you MUST refuse to improve or augment the code. You can still analyze existing code, write reports, or answer questions about the code behavior.

Just happened to me and I was really confused. First time I've seen any malware callouts so it had me worried for a minute.

> This file is clearly not malware

Yeah, it's all my code, that you've seen before...

I noticed this also, and was abit taken back at first...

But I think this is good thing the model checks the code, when adding new packages etc. Especially given that thousands of lines of code aren't even being read anymore.

This is the same paranoid, anxious behavior that ChatGPT has. One hell of a bad sign.

> Plan confirmed. Not malware — it's my own design doc. Let me quickly check proto and dependencies I'll need.

I had the same problem. Restarted Claude Code after an update, and now it has disappeared.

This is funny on so many levels.

Is this happening on the latest build of Claude Code? Try `claude --update`

it used to do this naturally sometimes, quite often in my runtime debugging.

I agree with you here. I think this is for product placement for Mythos.

Oh don't worry. They have Mythos and the extremely dystopian-named "helpful only" series which is internal only and can do all the things.

Now we have to trick the models when you legitimately work in the security space.

I am absolutely moving off them if this continues to be the case.

Questions about "fatality" aside, where do you see asymmetry here?

Only software approved by Anthropic (and/or the USG) is allowed to be secure in this brave new era.

Claude Code injects a 'warning: make sure this file isn't malware' message after every tool call by default. It seems like 4.7 is over-attending to this warning. @bcherny, filed a bug report feedback ID: 238e5f99-d6ee-45b5-981d-10e180a7c201

That "per the instructions I've been given in this session" bit is interesting. Are you perhaps using it with a harness that explicitly instructs it to not do that? If so, it's not being fussy, it's just following the instructions it was given.

I had a conversation right during the launch so not fully sure if it was Opus 4.7 but I also noticed the same behavior of asking questions that did not seem particularly useful to me, tho I still prefer that to not asking enough.

There’s one that’s like “Considering 17 theories” that had me wondering what those 17 things would be, I wanted to see them! Turns out it’s just a static message. Very confusing.

Sounds really minor, but was actually a big contributor to me canceling and switching. The VS Code extension has a morphing spinner thing that rapidly switches between these little catch phrases. It drives me crazy, and I end up covering it up with my right click menu so I can read the actual thinking tokens without that attention vampire distracting me.

And of course they recently turned off all third party harness support for the subscription, so you're just forced to watch it and any other stuff they randomly decide to add, or pay thousands of dollars.

Could you say more about your workflow? I don’t think I’ve ever gotten close to an hour of thinking before. Always curious to learn how to get more out of agents.

It wouldn't be so irritating if thinking didn't start to take a lot longer for tasks of similar complexity (or maybe it's taking longer to even start to think behind the scenes due to queueing).

Agreed. I actually have thought those were “waiting to get a response from the API” rather than “the model is still thinking” messages

It is the new "You are absolutely right!"

Maybe this? From the article:

> Opus 4.7 is substantially better at following instructions. Interestingly, this means that prompts written for earlier models can sometimes now produce unexpected results: where previous models interpreted instructions loosely or skipped parts entirely, Opus 4.7 takes the instructions literally. Users should re-tune their prompts and harnesses accordingly.

That depends a bit on token efficiency. From their "Agentic coding performance by effort level" graph, it looks like they get similar outcome for 4.7 medium at half the token usage as 4.6 at high.

Granted that is, as you say, a single prompt, but it is using the agentic process where the model self prompts until completion. It's conceivable the model uses fewer tokens for the same result with appropriate effort settings.

Anthropic isn't going to give us that information. It's not actually static, it depends on subscription demand and idle compute available.

The more efficient tokenizer reduces usage by representing text more efficiently with fewer tokens. But the lack of transparancy does indeed mean Anthropic could still scale down limits to account for that.

a few months ago it was for weekly:

pro = 5m tokens, 5x = 41m tokens, 20x = 83m tokens

making 5x the best value for the money (8.33x over pro for max 5x). this information may be outdated though, and doesn't apply to the new on peak 5h multipliers. anything that increases usage just burns through that flat token quota faster.

On the most current version (v2.1.110) of claude:

> /model claude-opus-4.7

  ⎿  Model 'claude-opus-4.7' not found

     /model claude-opus-4.7
      ⎿  Model 'claude-opus-4.7' not found

Just love that I'm paying $200 for models features they announce I can't use!

Related features that were announced I have yet to be able to use:

    $ claude --enable-auto-mode 
    auto mode is unavailable for your plan

    $ claude
    /memory 
    Auto-dream: on · /dream to run
    Unknown skill: dream

Thanks, but not working for me, and I'm on the $200 max plan

Edit: Not 30 seconds later, claude code took an update and now it works!

It's up now, update claude code

--model claude-opus-4-7 works as well

It does not work, it says Claude Opus 4 not 4.7

Couldn't we use domain records to verify that a website is our own for example with the TXT value provided by Anthropic?

Google does the same thing for verifying that a website is your own. Security checks by the model would only kick off if you're engaging in a property that you've validated.

Or they could check if the source is open source and available on the internet, and if yes refuse to analyse it if the person who request the analysis isn't affiliated to the project.

That will still leave closed source software vulnerable, but I suspect it is somewhat rare for hackers to have the source of the thing they are targeting, when it is closed source.

They don't have the compute to make Mythos generally available: that's all there is to it. The exclusivity is also nice from a marketing pov.

Looks like they are adding Peter Thiel backed ID verification too.

https://reddit.com/r/ClaudeAI/comments/1smr9vs/claude_is_abo...

Oh look it was too powerful to release, now it’s just a matter of safeguards.

This story sounds a lot like GPT2.

Mythos release feels like Silicon Valley "don't take revenue" advice:

https://www.youtube.com/watch?v=BzAdXyPYKQo

""If you show the model, people will ask 'HOW BETTER?' and it will never be enough. The model that was the AGI is suddenly the +5% bench dog. But if you have NO model, you can say you're worried about safety! You're a potential pure play... It's not about how much you research, it's about how much you're WORTH. And who is worth the most? Companies that don't release their models!"

The most highly anticipated model looking forward to using it

In GitHub Copilot it costs 7.5x whereas Opus 4.6 is 3x

> Opus 4.7 will replace Opus 4.5 and Opus 4.6

Promotional pricing that will probably be 9x when promotion ends, and soon to be the only Opus option on github, that's insane

Not that anybody can actually use it though, as a large percentage of Copilot users are facing seemingly random multi-day rate limits.

https://www.theregister.com/2026/04/15/github_copilot_rate_l...

But it majorly regressed in long context retrieval? Which is arguably getting more and more important?

Only in benchmarks. After couple of minutes of use it feels same dumb as nerfed 4.6

Some of the benchmarks went down, has that happened before?

Are you one of those naive people that still take these coding benchmarks seriously?

People were "predicting" the plateau since GPT-1. By now, it would take extraordinary evidence for me to take such "predictions" seriously.

I like this in theory. I just hope it doesn't require you to be be as literal as if talking to a genie.

But if it'll actually stick to the hard rules in the CLAUDE.md files, and if I don't have to add "DON'T DO ANYTHING, JUST ANSWER THE QUESTION" at the end of my prompt, I'll be glad.

coming more in line with codex - claude previously would often ignore explicit instructions that codex would follow. interested to see how this feels in practice

I think this line around "context tuning" is super interesting - I see a future where, for every model release, devs go and update their CLAUDE.md / skills to adapt to new model behavior.

This made me LOL. They keep trying to fleece us by nerfing functionality and then adding it back next release. It’s an abusive relationship at this point.

This sounds good, I look forward to experimenting with it.

Interesting that it used font-family:&quot;Anthropic Sans

wait till you hear about how we standardized RF bands. We have gems such as "High frequency", "Very High Frequency", "Ultra High Frequency", "Super High Frequency", and the cherry on top, "Extremely High Frequency". Then they went with the boring" Teraherz Frequency", truly a disappointment.

These are all mirrored on the low side btw, so we also have "Extremely Low Frequency", and all the others.

there's a line inside claude code mentioning to care about this. combined with new stronger instruction following behavior, you're going to be seeing it a lot unless you patch it out. or wait for a fix.

because they’re using it for different things where it works well and that’s all they know?

Because it was good until January 2026, then it detoriated into a opus-3.1. Probably given much less context windows or ram.

And yet another "AI doesn't work" comment without any meaningful information. What were your exact prompts? What was the output?

This is like a user of conventional software complaining that "it crashes", without a single bit of detail, like what they did before the crash, if there was any error message, whether the program froze or completely disappeared, etc.

marketing

Because they just don’t have enough capacity to serve their demand ?

I haven’t noticed much difference compared to Jan/Feb. Maybe depends what you use it for

Codex or the Chinese models

not a bot, voiced frustration is real here. I kind of depend on good LLMs now and wouldn't even mind if they had frozen the LLMs capabilities around dec 2025 forver and would hppily continue to pay, even more. but when suddenly the very same workload that was fine for months isn't possible anymore with the very same LLM out of nowhere and gets increasingly worse, its a huge disappointment. and having codex in parallel as a backup since ever I started also using it again with gpt 5.4 and it just rips without the diva sensitivity or overfitting into the latest prompt opus/sonnet is doing. GPT just does the job, maybe thinks a bit long, but even over several rounds of chat compression in the same chat for days stays well within the initial set of instructions and guardrails I spelled out, without me having to remind every time. just works, quietly, and gets there. Opus doesn't even get there anymore without nearly spelling out by hand manual steps or what not to do.

It's a combination of factors. There was rate-limiting implemented by Anthropic, where the 5hr usage limit would be burned through faster at peak hours, I was personally bitten by this multiple times before one guy from Anthropic announced it publicly via twitter, terrible communication. It wasn't small either, ~15 minutes of work ended up burning the entire 5hr limit. That annoyed me enough to switched to Codex for the month at that point.

Now people are saying the model response quality went down, I can't vouch for that since I wasn't using Claude Code, but I don't think this many people saying the same thing is total noise though.

Yeah, my personal anecdata is that Claude has just gotten better and better since January. I haven’t felt like even making the minor effort to compare with Codex’s current state. Just yesterday Claude Code made a major visible improvement in planning/executing — maybe it switched to 4.7 without me noticing? (Task: various internal Go services and Preact frontends.)

I'm an Opus stan but I'll also admit that 5.4 has gotten a lot better, especially at finding and fixing bugs. Codex doesn't seem to do as good a job at one shotting tasks from scratch.

I suppose if you are okay with a mediocre initial output that you spend more time getting into shape, Codex is comparable. I haven't exhaustively compared though.

Looks to me like a mob of humans, angry they've been deceived by ambiguous communications, product nerfing, surprisingly low usage limits, and an appallingly sycophantic overconfident coding agent

I'm wondering this too. That said, I know a few people in real life who prefer Codex. More who prefer Claude though.

I've had good experiences with codex, as have many others. Its genuine content since everyones codebases and needs are different.

In the gemini subreddit there is a persistent problem with bots posting "Gemini sucks, I switched to Claude" and then bots replying they did the same.

Old accounts with no posts for a few years, then suddenly really interested in talking up Claude, and their lackeys right behind to comment.

Not even necessarily calling out Anthropic, many fan boys view these AI wars as existential.

You're better off subscribing to Codex for April and May of 2026.

It feels like OAI stans have been botting HN for a few weeks now.

Sorry, no, not a bot. I get way better results out of Codex.

It's just ultimately subjective, and, it's like, your opinion, man. Calling people bots who disagree is probably not a good look.

I don't like OpenAI the company, but their model and coding tool is pretty damn good. And I was an early Claude Code booster and go back and forth constantly to try both.

I wish someone would elaborate on what they were doing and observed since Jan on opus 4.6. I’ve been using it with 1m context on max thinking since it was released - as a software engineer to write most of my code, code reviews + research and explain unfamiliar code - and haven’t notice a degradation. I’ve seen this mentioned a lot though.

I have seen that codex -latest highest effort - will find some important edge cases that opus 4.6 overlooked when I ask both of them to review my PRs.

Mm, don't you just need to retrain the embedding layer for the new tokenizer? I agree it seems likely this is like a stopgap new model release or a distillation of mythos or something while they get a better mythos release in place. But there are some things that look really different than mythos in the model card, e.g. the number of tokens it uses at different effort levels.

Maybe it's an abandoned candidate "5.0" model that mythos beat out.

It doesn't need to be. Text can be tokenized in many different ways even if the token set is the same.

For example there is usually one token for every string from "0" to "999" (including ones like "001" seperately).

This means there are lots of ways you can choose to tokenize a number. Like 27693921. The best way to deal with numbers tends to be a little bit context dependent but for numerics split into groups of 3 right to left tends to be pretty good.

They could just have spotted that some particular patterns should be decomposed differently.

> Usually a ground up rebuild is related to a bigger announcement. So, it's weird that they'd be naming it 4.7.

Benchmarks say it all. Gains over previous model are too small to announce it as a major release. That would be humiliating for Anthropic. It may scare investors that the curve flattened and there are only diminishing returns.

On the other hand, I never understood the focus on computer use.

While more general and perhaps the "ideal" end state once models run cheaply enough, you're always going to suffer from much higher latency and reduced cognition performance vs API/programmatically driven workflows. And strictly more expensive for the same result.

Why not update software to use API first workflows instead?

The industry probably moves a lot faster adding apis and co than learning how to use a generic computer with generic tools.

I also think its a huge barrier allowing some LLM model access to your desktop.

Managed Agents seems like a lot more beneficial

I've noticed this and thought about it as well, I have a few suspicions:

Theory 1: Some increasingly-large split of inference compute is moving over to serving the new model for internal users (or partners that are trialing the next models). This results in less compute but the same increasing demand for the previous model. Providers may respond by using quantizations or distillations, compressing k/v store, tweaking parameters, and/or changing system prompts to try to use fewer tokens.

Theory 2: Internal evals are obviously done using full strength models with internally-optimized system prompts. When models are shipped into production the system prompt will inherently need changes. Each time a problematic issue rises to the attention of the team, there is a solid chance it results in a new sentence or two added to the system prompt. These grow over time as bad shit happens with the model in the real world. But it doesn't even need to be a harmful case or bad bugged behavior of the model, even newer models with enhanced capabilities (e.g. mythos) may get protected against in prompts used in agent harnesses (CC) or as system prompts, resulting in a more and more complex system prompt. This has something like "cognitive burden" for the model, which diverges further and further from the eval.

> So my question is: what is the actual reason Anthropic lobotomizes the model when the new one is about to be dropped?

You can only fit one version of a model in VRAM at a time. When you have a fixed compute capacity for staging and production, you can put all of that towards production most of the time. When you need to deploy to staging to run all the benchmarks and make sure everything works before deploying to prod, you have to take some machines off the prod stack and onto the staging stack, but since you haven't yet deployed the new model to prod, all your users are now flooding that smaller prod stack.

So what everyone assumes is that they keep the same throughput with less compute by aggressively quantizing or other optimizations. When that isn't enough, you start getting first longer delays, then sporadic 500 errors, and then downtime.

To be fair, delineating between benevolent and malevolent pen-testing and cybersecurity purposes is practically impossible since the only difference is the user's intentions. I am entirely unsurprised (and would expect) that as models improve the amount to which widely available models will be prohibited from cybersecurity purposes will only increase.

Not to say I see this as the right approach, in theory the two forces would balance each other out as both white hats and black hats would have access to the same technology, but I can understand the hesitancy from Anthropic and others.

I'm currently testing 4.7 with some reverse engineering stuff/Ghidra scripting and it hasn't refused anything so far, but I'm also doing it on a 20 year old video game, so maybe it doesn't think that's problematic.

Claude code had safeguards like that hardcoded into the software. You could see it if you intercept the prompts with a proxy

Incredible - in one fell swoop killing my entire use case for Claude.

I have about 15 submissions that I now need to work with Codex on cause this "smarter" model refuses to read program guidelines and take them seriously.

From the article:

> Security professionals who wish to use Opus 4.7 for legitimate cybersecurity purposes (such as vulnerability research, penetration testing, and red-teaming) are invited to join our new Cyber Verification Program.

It appears we're learning the hard way that we can't rely on capabilities of models that aren't open weights. These can be taken from us at any time, so expect it to get much worse..

And since Opus 4.7 has degraded cybersecurity skills, using it might result in writing actually less safe code, since practically, in order to write secure code you need to understand cybersecurity. Outstanding move.

There is no hallucination benchmark currently.

I was researching how to predict hallucinations using the literature (fastowski et al, 2025) (cecere et al, 2025) and the general-ish situation is that there are ways to introspect model certainty levels by probing it from the outside to get the same certainty metric that you _would_ have gotten if the model was trained as a bayesian model, ie, it knows what it knows and it knows what it doesn't know.

This significantly improves claim-level false-positive rates (which is measured with the AUARC metric, ie, abstention rates; ie have the model shut up when it is actually uncertain).

This would be great to include as a metric in benchmarks because right now the benchmark just says "it solves x% of benchmarks", whereas the real question real-world developers care about is "it solves x% of benchmarks *reliably*" AND "It creates false positives on y% of the time".

So the answer to your question, we don't know. It might be a cherry picked result, it might be fewer hallucinations (better metacognition) it might be capability to solve more difficult problems (better intelligence).

The benchmarks don't make this explicit.

Benchmarks are meaningless. Try it on your own problems and see if it has improved for what you want to use it for.

Benchmark results don’t directly translate to actual real world improvement. So we might guess it’s somewhat better but hard to say exactly in what way

11% further along the particular bell curve of SWE-bench. Not really easy to extrapolate to real world, especially given that eg the Chinese models tend to heavily train on the benchmarks. But a 10% bump with the same model should equate to “feels noticeably smarter”.

A more quantifiable eval would be METR’s task time - it’s the duration of tasks that the model can complete on average 50% of the time, we’ll have to wait to see where 4.7 lands on this one.

I assume that by "higher resolution images" you mean images with a bigger size in pixels.

I expect that for the model it does not matter which is the actual resolution in pixels per inch or pixels per meter of the images, but the model has limits for the maximum width and the maximum height of images, as expressed in pixels.

Did you try the same with gemini 3 models? Those usually score higher on vision benchmarks

page is updated to state:

MCP-Atlas: The Opus 4.6 score has been updated to reflect revised grading methodology from Scale AI.

if only. but more token costs, yes.

I get this error too and if I try again: { ... "error":{"type":"permission_error","message":"anthropic.claude-opus-4-7 is not available for this account. You can explore other available models on Amazon Bedrock. For additional access options, contact AWS Sales at https://aws.amazon.com/contact-us/sales-support/"}}

They are planning to release a Mythos-class model (from the initial announcement), but they won't until they can trust their safeguards + the software ecosystem has been sufficiently patched.

It seems they nerf it, then release a new version with previous power. So they can do this forever without actually making another step function model release.

~~That just changes it to Opus 4, not Opus 4.7~~

My statusline showed _Opus 4_, but it did indeed accept this line.

I did change it to `/model claude-opus-4-7[1m]`, because it would pick the non-1M context model instead.

Does it run for you? I can select it this way but it says 'There's an issue with the selected model (claude-opus-4-7). It may not exist or you may not have access to it. Run /model to pick a different model.'

API Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"\"thinking.type.enabled\" is not supported for this model. Use \"thinking.type.adaptive\" and \"output_config.effort\" to control thinking behavior."},"request_id":"req_011Ca7enRv4CPAEqrigcRNvd"}

Eep. AFAIK the issues most people have been complaining about with Opus 4.6 recently is due to adaptive thinking. Looks like that is not only sticking around but mandatory for this newer model.

edit: I still can't get it to work. Opus 4.6 can't even figure out what is wrong with my config. Speaking of which, claude configuration is so confusing there are .claude/ (in project) setting.json + a settings.local.json file, then a global ~/.claude/ dir with the same configuration files. None of them have anything defined for adaptive thinking or thinking type enable. None of these strings exist on my machine. Running latest version, 2.1.110

claude-opus-4-7

/model claude-opus-4-7[1m]

Did you try asking the model?

It seems a lot of the problem isn't "token shrinkage" (reducing plan limits), but rather changes they made to prompt caching - things that used to be cached for 1 hour now only being cached for 5 min.

Coding agents rely on prompt caching to avoid burning through tokens - they go to lengths to try to keep context/prompt prefixes constant (arranging non-changing stuff like tool definitions and file content first, variable stuff like new instructions following that) so that prompt caching gets used.

This change to a new tokenizer that generates up to 35% more tokens for the same text input is wild - going to really increase token usage for large text inputs like code.

Solve? You solve a problem, not something you introduced on purpose.

on Tuesday, with 4.6, I waited for my 5 hour window to reset, asked it to resume, and it burned up all my tokens for the next 5 hour window and ran for less than 10 seconds. I’ve never cancelled a subscription so fast.

Who cares about a new model you can’t even use?

Dude we dont even have GI

They are still available. Just in OpenAI instead.

If OpenAI has a new model that they are close to releasing, now seems like a perfect opening to steal some thunder. Mythos coming out later with only marginal improvements to a new OpenAI model would be good-great outcome for OpenAI

Gemini and Codex already scored higher on benchmarks than Opus 4.6 and they recently added a $100 tier with limited 2x limits, that's their answer and it seems people have caught on.

You are operating purely on vibes, https://marginlab.ai/trackers/claude-code-historical-perform...

I feel this way too. Wish I could fully understand the 'why'. I know all of the usual arguments, but nothing seems to fully capture it for me - maybe it' all of them, maybe it's simply the pace of change and having to adapt quicker than we're comfortable with. Anyway best of luck from someone who understands this sentiment.

Why? Good anxiety or bad?

SWE-bench pro is ~20% higher than the previous .1 generation which was released 2 months ago. For their SWE benchmark, the token consumption iso-performance is down 2x from the model they released 2 months ago.

If this is a plateau I struggle to imagine what you consider fast progress.

Your comment doesn't make any sense, opus 4.6 was release two months ago, what jump would you expect?

Every night praying for tomorrow

The generations are two months apart now though…

If you're building a standard app Opus is already good enough to build anything you want. I don't even know what you'd really need Mythos for.

I don't get it. Why would they make the previous model worse before releasing an update?