Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Claude Opus 4.7

anthropic.com

Post reply on HN
Print view View original

781–790 of 1001 posts

Re: Claude Opus 4.7

#781

Post by buf »

ed_mercer wrote: post #713
andai wrote: post #679

Earlier quoted context omitted.

Well, I can share my experience from a few days ago. Gave the same task (a major refactor) to both Claude and Codex. Codex finished in 5 minutes, Claude was still spinning after 20 minutes. Also it used up all my usage, about twice over (the 5-hour window rolled over in the middle of the task, so the usage for one task added up to 192%). Codex usage was 9%. So, 21x difference there, lol They're saying there's bugs la…

> Claude was still spinning after 20 minutes. I have been using Claude Code on a medium codebase (~2000 files, ~1M lines of code) for over a year and have never had to wait this long. Also I'm on the max plan and have not seen these limits at all.

Just chipping in to say that I've never seen it churn for more than 20 minutes in two years worth of usage. The longest I've ever seen it churn is when I had it give extremely detailed analysis of five fictional novels simultaneously.

Re: Claude Opus 4.7

#782

Post by devmor »

simonw wrote: post #203

I'm finding the "adaptive thinking" thing very confusing, especially having written code against the previous thinking budget / thinking effort / etc modes: https://platform.claude.com/docs/en/build-with-claude/adapti... 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: https://platform.claude.com/docs/en/build-…

> 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

That’s extremely bothersome because half of what helps teams build better guardrails and guidelines for agents is the ability to do deep analysis on session transcripts.

I guess we shouldn’t be surprised these vendors want to do everything they can to force users to rely explicitly on their offerings.

Re: Claude Opus 4.7

#783

Post by raylad »

sothatsit wrote: post #759
raylad wrote: post #751

I am using 4.7 with the default extra high thinking, and it is clearly very stupid. It's worse than old Sonnet 4.5. I had it suggest some parameters for BCFtools and it suggested parameters that would do the opposite of what I wanted to do. I pointed out the error and it apologized. It also is not taking any initiative to check things, but wants me to check them (ie: file contents, etc.). And it is claiming that thin…

You can’t make up your mind about a model by using it on one task. Especially to say it’s such a bad downgrade after that is ludicrous. I’ve had great experiences with it this morning.

That was more than one task. It was 3.

I also had Opus 4.7 and Opus 4.6 do audits of a very long document using identical prompts. I then had Codex 5.4 compare the audits. Codex found that 4.6 did a far better job and 4.7 had missed things and added spurious information.

I then asked a new session of Opus 4.7 if it agreed or disagreed with the Codex audit and it agreed with it.

I also agreed with it.

Re: Claude Opus 4.7

#784

Post by vbezhenar »

turblety wrote: post #296
corlinp wrote: post #194

I'm running it for the first time and this is what the thinking looks like. Opus seems highly concerned about whether or not I'm asking it to develop malware. > This is _, not malware. Continuing the brainstorming process. > Not malware — standard _ code. Continuing exploration. > Not malware. Let me check front-end components for _. > Not malware. Checking validation code and _. > Not malware. > Not malware.

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...

"generating fake news, impersonating people, or automating abusive or spam comments on social media"

So it seems that these fears were founded. Doesn't seem to be a "theatre".

Re: Claude Opus 4.7

#785

Post by vdalhambra »

simonw wrote: post #203

I'm finding the "adaptive thinking" thing very confusing, especially having written code against the previous thinking budget / thinking effort / etc modes: https://platform.claude.com/docs/en/build-with-claude/adapti... 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: https://platform.claude.com/docs/en/build-…

[flagged]

Re: Claude Opus 4.7

#786

Post by sagarpatil »

aurareturn wrote: post #31
endymion-light wrote: post #16

I'm not sure how much I trust Anthropic recently. 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. Anthropic need to build back some trust and communicate throtelling/reasoning caps more clearly.

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.

It worked. Although I have a Claude Code subscription, I got the ChatGPT Pro plan, and 5.4 xHigh at 1.5x speed was better than 4.6 with adaptive thinking disabled. I was working all day, about 8 hours, and did not run into any limits. 5.4 surprised me many times by doing things I usually would not do myself, because I am lazy, so yeah, I am sticking with 5.4 for now until all the Claude drama is over.

Re: Claude Opus 4.7

#787

Post by sagarpatil »

christophilus wrote: post #757
Jcampuzano2 wrote: post #529

Earlier quoted context omitted.

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.…

Same. Codex is faster and more consistent in the last few weeks for me vs Claude Code. I also don’t hit limits anywhere near as frequently.

Same. I’ve lived in Claude Code since the beta release and last couple of weeks was horrible. I’ve been using codex for last couple of days and it’s much smarter than 4.6.

Re: Claude Opus 4.7

#788

Post by dbdr »

ang_cire wrote: post #607
alaudet wrote: post #600

Serious question about using Claude for coding. I maintain a couple of small opensource applications written in python that I created back in 2014/2015. I have used Claude Code to improve one of my projects with features I have wanted for a long time but never really had the time to do. The only way I felt comfortable using Claude Code was holding its hand through every step, doing test driven changes and manually re…

> Humans are lazy though and I can't help but feel we are being inundated with sketchy apps doing all kinds of things the authors don't even understand... there is a good chance they have no clue what they created. I have bad news for you about the executives and salespeople who manage and sell fully-human-coded enterprise software (and about the actual quality of much of that software)... I think people who aren't w…

> not doing the proper reviews, but once again, this is not remotely unique to AI code; this is what 99% of companies are already doing.

But is the scale similar, or will AI coding make the problem significantly worse?

Re: Claude Opus 4.7

#789

Post by YZF »

drewnick wrote: post #233
lanyard-textile wrote: post #200

This comment thread is a good learner for founders; look at how much anguish can be put to bed with just a little honest communication. 1. Oops, we're oversubscribed. 2. Oops, adaptive reasoning landed poorly / we have to do it for capacity reasons. 3. Here's how subscriptions work. Am I really writing this bullet point? As someone with a production application pinned on Opus 4.5, it is extremely difficult to tell ap…

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

For me 4.6 has been a noticeable leap in performance from 4.5. I'm not missing 4.5 at all.

Re: Claude Opus 4.7

#790

Post by mrifaki »

the adaptive thinking complaints in this thread are interesting because they are basically the same verifier quality problem showing up in a different costume the model has to decide how hard to think before knowing how hard the problem is and that meta decision is itself a hard problem that nobody has solved cleanly not in RL not in speculative decoding not in branch prediction, the fact that disabling adaptive thinking and forcing high effort restores quality tells us the router is underthinning not that the model got worse which means anthropic is trading user experience for compute savings whether or not they frame it that way
Post reply on HN