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

Claude Sonnet 4.6

anthropic.com

Post reply on HN
Print view View original

341–350 of 1001 posts

Re: Claude Sonnet 4.6

#341

Post by wongarsu »

andrewchilds wrote: post #244

Many people have reported Opus 4.6 is a step back from Opus 4.5 - that 4.6 is consuming 5-10x as many tokens as 4.5 to accomplish the same task: https://github.com/anthropics/claude-code/issues/23706 I haven't seen a response from the Anthropic team about it. I can't help but look at Sonnet 4.6 in the same light, and want to stick with 4.5 across the board until this issue is acknowledged and resolved.

Keep in mind that the people who experience issues will always be the loudest.

I've overall enjoyed 4.6. On many easy things it thinks less than 4.5, leading to snappier feedback. And 4.6 seems much more comfortable calling tools: it's much more proactive about looking at the git history to understand the history of a bug or feature, or about looking at online documentation for APIs and packages.

A recent claude code update explicitly offered me the option to change the reasoning level from high to medium, and for many people that seems to help with the overthinking. But for my tasks and medium-sized code bases (far beyond hobby but far below legacy enterprise) I've been very happy with the default setting. Or maybe it's about the prompting style, hard to say

Re: Claude Sonnet 4.6

#342

Post by Topfi »

andrewchilds wrote: post #244

Many people have reported Opus 4.6 is a step back from Opus 4.5 - that 4.6 is consuming 5-10x as many tokens as 4.5 to accomplish the same task: https://github.com/anthropics/claude-code/issues/23706 I haven't seen a response from the Anthropic team about it. I can't help but look at Sonnet 4.6 in the same light, and want to stick with 4.5 across the board until this issue is acknowledged and resolved.

In my evals, I was able to rather reliably reproduce an increase in output token amount of roughly 15-45% compared to 4.5, but in large part this was limited to task inference and task evaluation benchmarks. These are made up of prompts that I intentionally designed to be less then optimal, either lacking crucial information (requiring a model to output an inference to accomplish the main request) or including a request for a less than optimal or incorrect approach to resolving a task (testing whether and how a prompt is evaluated by a model against pure task adherence). The clarifying question many agentic harnesses try to provide (with mixed success) are a practical example of both capabilities and something I do rate highly in models, as long as task adherence isn't affected overly negatively because of it.

In either case, there has been an increase between 4.1 and 4.5, as well as now another jump with the release of 4.6. As mentioned, I haven't seen a 5x or 10x increase, a bit below 50% for the same task was the maximum I saw and in general, of more opaque input or when a better approach is possible, I do think using more tokens for a better overall result is the right approach.

In tasks which are well authored and do not contain such deficiencies, I have seen no significant difference in either direction in terms of pure token output numbers. However, with models being what they are and past, hard to reproduce regressions/output quality differences, that additionally only affected a specific subset of users, I cannot make a solid determination.

Regarding Sonnet 4.6, what I noticed is that the reasoning tokens are very different compared to any prior Anthropic models. They start out far more structured, but then consistently turn more verbose akin to a Google model.

Re: Claude Sonnet 4.6

#343

Post by iamdelirium »

anonym29 wrote: post #243
fullstackchris wrote: post #225

Earlier quoted context omitted.

google is "evil" ok buddy i mean what clown show are we living in at this point - claims like this simply running rampant with 0 support or references

They literally removed "don't be evil" from their internal code of conduct. That wasn't even a real binding constraint, it was simply a social signalling mechanism. They aren't even willing to uphold the symbolic social fiction of not being evil. https://en.wikipedia.org/wiki/Don't_be_evil Google, like Microsoft, Apple, Amazon, etc were, and still are, proud partners of the US intelligence community. That same US IC…

Don't be evil was never removed. It was just moved to the bottom.

https://abc.xyz/investor/board-and-governance/google-code-of...

Re: Claude Sonnet 4.6

#344

Post by Pearse »

chrisweekly wrote: post #51
eth0up wrote: post #40

Earlier quoted context omitted.

I am casually 'researching' this in my own, disorderly way. But I've achieved repeatable results, mostly with gpt for which I analyze its tendency to employ deflective, evasive and deceptive tactics under scrutiny. Very very DARVO. Being just sum guy, and not in the industry, should I share my findings? I find it utterly fascinating, the extent to which it will go, the sophisticated plausible deniability, and the dis…

DARVO stands for "Deny, Attack, Reverse Victim and Offender," and it is a manipulation tactic often used by perpetrators of wrongdoing, such as abusers, to avoid accountability. This strategy involves denying the abuse, attacking the accuser, and claiming to be the victim in the situation.

Thanks for the context

Re: Claude Sonnet 4.6

#345

Post by bredren »

KronisLV wrote: post #335
MrCheeze wrote: post #321

Earlier quoted context omitted.

In my experience with the models (watching Claude play Pokemon), the models are similar in intelligence, but are very different in how they approach problems: Opus 4.5 hyperfocuses on completing its original plan, far more than any older or newer version of Claude. Opus 4.6 gets bored quickly and is constantly changing its approach if it doesn't get results fast. This makes it waste more time on"easy" tasks where the…

I got the Max subscription and have been using Opus 4.6 since, the model is way above pretty much everything else I've tried for dev work and while I'd love for Anthropic to let me (easily) work on making a hostable server-side solution for parallel tasks without having to go the API key route and not have to pay per token, I will say that the Claude Code desktop app (more convenient than the TUI one) gets me most of…

Can you explain what you mean by your parallel tasks limitation?

Re: Claude Sonnet 4.6

#346

Post by kingofthehill98 »

retinaros wrote: post #122
energy123 wrote: post #118

Earlier quoted context omitted.

Grok usage is the most mystifying to me. Their model isn't in the top 3 and they have bad ethics. Like why would anyone bother for work tasks.

The X grok feature is one of the best end user feature or large scale genai

What?! That's well regarded as one of the worst features introduced after the Twitter acquisition.

Any thread these days is filled with "@grok is this true?" low effort comments. Not to mention the episode in which people spent two weeks using Grok to undress underage girls.

Re: Claude Sonnet 4.6

#347

Post by alkhatib »

KronisLV wrote: post #335
MrCheeze wrote: post #321

Earlier quoted context omitted.

In my experience with the models (watching Claude play Pokemon), the models are similar in intelligence, but are very different in how they approach problems: Opus 4.5 hyperfocuses on completing its original plan, far more than any older or newer version of Claude. Opus 4.6 gets bored quickly and is constantly changing its approach if it doesn't get results fast. This makes it waste more time on"easy" tasks where the…

I got the Max subscription and have been using Opus 4.6 since, the model is way above pretty much everything else I've tried for dev work and while I'd love for Anthropic to let me (easily) work on making a hostable server-side solution for parallel tasks without having to go the API key route and not have to pay per token, I will say that the Claude Code desktop app (more convenient than the TUI one) gets me most of…

Try https://conductor.build

I started using it last week and it’s been great. Uses git worktrees, experimental feature (spotlight) allows you to quickly check changes from different agents.

I hope the Claude app will add similar features soon

Re: Claude Sonnet 4.6

#348

Post by Someone1234 »

simianparrot wrote: post #339

How do people keep track of all these versions and releases of all these models and their pros/cons? Seems like a fulltime hobby to me. I'd rather just improve my own skills with all that time and energy

Unless you're interested in this type of stuff, I'm not sure you really need to. Claude, Google, and ChatGPT have been fairly aggressive at pushing you towards whatever their latest shiny is and retiring the old one.

Only time it matters if you're using some type of agnostic "router" service.

Re: Claude Sonnet 4.6

#349

Post by bearjaws »

krystofee wrote: post #337

Does anyone know when will possibly arrive 1M context windows to at least MAX x20 subscriptions for claude code? I would even pay x50 if it allowed that. API usage is too expensive.

Based on their API pricing a 1M context plan should be 2x the price roughly.

My bets are its more the increased hardware demand that they don't want to deal with currently.

Re: Claude Sonnet 4.6

#350

Post by srvo »

andsoitis wrote: post #77

I’m voting with my dollars by having cancelled my ChatGPT subscription and instead subscribing to Claude. Google needs stiff competition and OpenAI isn’t the camp I’m willing to trust. Neither is Grok. I’m glad Anthropic’s work is at the forefront and they appear, at least in my estimation, to have the strongest ethics.

Ethics often fold under the face of commercial pressure.

The pentagon is thinking [1] about severing ties with anthropic because of its terms of use, and in every prior case we've reviewed (I'm the Chief Investment Officer of Ethical Capital), the ethics policy was deleted or rolled back when that happens.

Corporate strategy is (by definition) a set of tradeoffs: things you do, and things you don't do. When google (or Microsoft, or whoever) rolls back an ethics policy under pressure like this, what they reveal is that ethical governance was a nice-to-have, not a core part of their strategy.

We're happy users of Claude for similar reasons (perception that Anthropic has a better handle on ethics), but companies always find new and exciting ways to disappoint you. I really hope that anthropic holds fast, and can serve in future as a case in point that the Public Benefit Corporation is not a purely aesthetic form.

But you know, we'll see.

[1] https://thehill.com/policy/defense/5740369-pentagon-anthropi...

Post reply on HN