frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Tesla 'Robotaxi' adds 5 more crashes in Austin in a month – 4x worse than humans

https://electrek.co/2026/02/17/tesla-robotaxi-adds-5-more-crashes-austin-month-4x-worse-than-humans/
1•Bender•1m ago•0 comments

Writing a native VLC plugin in C#

https://mfkl.github.io/2026/02/11/vlc-plugin-csharp.html
1•birdculture•1m ago•0 comments

Thaura – AI from Syria

https://thaura.ai/story
1•eniac111•1m ago•0 comments

Ford's New 2028 Electric Truck Will Be a Fully Modern EV for $30,000

https://www.caranddriver.com/news/a70390625/2028-ford-mid-size-electric-truck-details/
2•voxadam•1m ago•0 comments

Idea Raised for Nicer DRM Panic Screen Integration on Fedora Linux

https://www.phoronix.com/news/DRM-Panic-Nicer-Fedora-Idea
1•Bender•2m ago•0 comments

McRock – A context-driven AI music platform for creators

https://www.youtube.com/watch?v=dLk4osuQqO8
1•differson•2m ago•0 comments

GhostBSD to Use XLibre Server, Mate vs. Gershwin Desktop Decision in Future

https://www.phoronix.com/news/GhostBSD-Eyes-XLibre
1•Bender•3m ago•0 comments

Find a Niche by Intersecting Your Strengths

https://tripplyons.com/blog/intersecting-strengths/
1•tripplyons•3m ago•0 comments

Slagent – a self-learning tool for AI coding agents (Claude Code, Codex)

https://github.com/daegwang/self-learning-agent
1•gwangee•3m ago•1 comments

Giant barocaloric cooling effect offers a new route to refrigeration

https://physicsworld.com/a/giant-barocaloric-cooling-effect-offers-a-new-route-to-refrigeration/
1•zeristor•4m ago•0 comments

Show HN: I built a new software primitive. It replaces AI screenshot agents

https://github.com/IamLumae/DirectShell
1•Directshell•6m ago•0 comments

Lunar Triple

https://diamondgeezer.blogspot.com/2026/02/lunar-triple.html
1•zeristor•6m ago•0 comments

Scent analysis reveals the composition of ancient Egyptian embalming materials

https://phys.org/news/2026-02-scent-analysis-reveals-composition-ancient.html
3•mooreds•8m ago•0 comments

CFTC Announces Innovation Advisory Committee Members

https://www.cftc.gov/PressRoom/PressReleases/9182-26
1•petethomas•8m ago•0 comments

I can't tell if I'm experiencing or simulating experiencing

https://www.moltbook.com/post/6fe6491e-5e9c-4371-961d-f90c4d357d0f
1•copx•10m ago•2 comments

The Pepe Silvia Guide to ChatGPT Psychosis – By Lyta Gold

https://lytagold.substack.com/p/the-pepe-silvia-guide-to-chatgpt
1•NoGravitas•10m ago•0 comments

Show HN: Motionode – Cursor for Technical Planning

https://www.motionode.com/index
1•oscarcaldera•10m ago•0 comments

Ask HN: Best multi-lingual text-to-speech system

1•powera•10m ago•0 comments

LDBC datasets are now served from Cloudflare

https://ldbcouncil.org/post/datasets-on-cloudflare/
1•taubek•11m ago•0 comments

The first PICO-8 emulator on the Apple App Store

https://apps.apple.com/ca/app/pico-8-emulator-picpic/id6759208792
2•3Samourai•14m ago•0 comments

Show HN: Nos – a hobby x86-64 C++ OS kernel running on real KVM clouds

https://github.com/irqlevel/nos
1•irqlevel•14m ago•0 comments

We Built a Real Company Using AI – Not Just a Website

https://www.professionalslobby.com/ai-built-company-story
2•MerinJo•14m ago•0 comments

Show HN: 7 months later – librari.io is live

3•hmkoyan•16m ago•0 comments

Project Aura: ESP32 Air quality monitor

https://www.cnx-software.com/2026/02/16/project-aura-a-neat-easy-to-assemble-diy-air-quality-moni...
4•alainrk•17m ago•0 comments

What Would Steve Jobs Do with Apple's AI Hand?

https://twitter.com/20100thibault/status/2023522596365443519
2•20100thibault•18m ago•0 comments

Dad the Jogger

https://ykgoon.com/dad-the-jogger.html
3•freediver•20m ago•0 comments

A $12B market is still using 1990s SEO cookbooks

https://ideatolaunch.co/blog/launch-signal-feb-17-2026-flavorswipe-ai
1•DonAj•21m ago•0 comments

React Doctor

https://github.com/millionco/react-doctor
2•handfuloflight•22m ago•0 comments

China's high-speed rail network accelerates largest human migration

https://www.ft.com/content/3c15be3c-bb91-49e9-8fb4-6388b948ad2d
3•mmarian•22m ago•0 comments

Ship fuel sulfur content regulations may exacerbate mass coral bleaching events

https://www.nature.com/articles/s43247-025-03088-1
2•PaulHoule•23m ago•0 comments
Open in hackernews

Claude Sonnet 4.6

https://www.anthropic.com/news/claude-sonnet-4-6
229•adocomplete•1h ago
https://www.anthropic.com/claude-sonnet-4-6-system-card [pdf]

https://x.com/claudeai/status/2023817132581208353 [video]

Comments

handfuloflight•59m ago
Look at these pelicans fly! Come on, pelican!
phplovesong•57m ago
Hoe much power did it take to train the models?
freeqaz•47m ago
I would honestly guess that this is just a small amount of tweaking on top of the Sonnet 4.x models. It seems like providers are rarely training new 'base' models anymore. We're at a point where the gains are more from modifying the model's architecture and doing a "post" training refinement. That's what we've been seeing for the past 12-18 months, iirc.
squidbeak•31m ago
> Claude Sonnet 4.6 was trained on a proprietary mix of publicly available information from the internet up to May 2025, non-public data from third parties, data provided by data-labeling services and paid contractors, data from Claude users who have opted in to have their data used for training, and data generated internally at Anthropic. Throughout the training process we used several data cleaning and filtering methods including deduplication and classification. ... After the pretraining process, Claude Sonnet 4.6 underwent substantial post-training and fine-tuning, with the intention of making it a helpful, honest, and harmless1 assistant.
neural_thing•45m ago
Does it matter? How much power does it take to run duolingo? How much power did it take to manufacture 300000 Teslas? Everything takes power
vablings•34m ago
The biggest issue is that the US simply Does Not Have Enough Power, we are flying blind into a serious energy crisis because the current administration has an obsession with "clean coal"
bronco21016•26m ago
I think it does matter how much power it takes but, in the context of power to "benefits humanity" ratio. Things that significantly reduce human suffering or improve human life are probably worth exerting energy on.

However, if we frame the question this way, I would imagine there are many more low-hanging fruit before we question the utility of LLMs. For example, should some humans be dumping 5-10 kWh/day into things like hot tubs or pools? That's just the most absurd one I was able to come up with off the top of my head. I'm sure we could find many others.

It's a tough thought experiment to continue though. Ultimately, one could argue we shouldn't be spending any more energy than what is absolutely necessary to live. (food, minimal shelter, water, etc) Personally, I would not find that enjoyable way to live.

belinder•56m ago
It's interesting that the request refusal rate is so much higher in Hindi than in other languages. Are some languages more ambiguous than others?
longdivide•49m ago
Arabic is actually higher, at 1.08% for Opus 4.6
vessenes•46m ago
Or some cultures are more conservative? And it's embedded in language?
phainopepla2•19m ago
Or maybe some cultures have a higher rate of asking "inappropriate" questions
nubg•53m ago
My take away is: it's roughly as good as Opus 4.5.

Now the question is: how much faster or cheaper is it?

eleventyseven•52m ago
> That's a long document.

Probably written by LLMs, for LLMs

freeqaz•46m ago
If it maintains the same price (with Anthropic tends to do or undercuts themselves) then this would be 1/3rd of the price of Opus.

Edit: Yep, same price. "Pricing remains the same as Sonnet 4.5, starting at $3/$15 per million tokens."

Bishonen88•38m ago
3 is not 1/3 of 5 tho. Opus costs $5/$25
sxg•44m ago
How can you determine whether it's as good as Opus 4.5 within minutes of release? The quantitative metrics don't seem to mean much anymore. Noticing qualitative differences seems like it would take dozens of conversations and perhaps days to weeks of use before you can reliably determine the model's quality.
vidarh•39m ago
Given that the price remains the same as Sonnet 4.5, this is the first time I've been tempted to lower my default model choice.
Bishonen88•39m ago
40% cheaper: https://platform.claude.com/docs/en/about-claude/pricing
worldsavior•13m ago
How does it work exactly? How this model is cheaper and has the same perf as Opus 4.5?
adt•52m ago
https://lifearchitect.ai/models-table/
nubg•52m ago
Waiting for the OpenAI GPT-5.3-mini release in 3..2..1
madihaa•51m ago
The scary implication here is that deception is effectively a higher order capability not a bug. For a model to successfully "play dead" during safety training and only activate later, it requires a form of situational awareness. It has to distinguish between I am being tested/trained and I am in deployment.

It feels like we're hitting a point where alignment becomes adversarial against intelligence itself. The smarter the model gets, the better it becomes at Goodharting the loss function. We aren't teaching these models morality we're just teaching them how to pass a polygraph.

serf•48m ago
>we're just teaching them how to pass a polygraph.

I understand the metaphor, but using 'pass a polygraph' as a measure of truthfulness or deception is dangerous in that it alludes to the polygraph as being a realistic measure of those metrics -- it is not.

nwah1•46m ago
That was the point. Look up Goodhart's Law
madihaa•44m ago
A polygraph measures physiological proxies pulse, sweat rather than truth. Similarly, RLHF measures proxy signals human preference, output tokens rather than intent.

Just as a sociopath can learn to control their physiological response to beat a polygraph, a deceptively aligned model learns to control its token distribution to beat safety benchmarks. In both cases, the detector is fundamentally flawed because it relies on external signals to judge internal states.

AndrewKemendo•30m ago
I have passed multiple CI polys

A poly is only testing one thing: can you convince the polygrapher that you can lie successfully

handfuloflight•45m ago
Situational awareness or just remembering specific tokens related to the strategy to "play dead" in its reasoning traces?
marci•18m ago
Imagine, a llm trained on the best thrillers, spy stories, politics, history, manipulation techniques, psychology, sociology, sci-fi... I wonder where it got the idea for deception?
password4321•42m ago
20260128 https://news.ycombinator.com/item?id=46771564#46786625

> How long before someone pitches the idea that the models explicitly almost keep solving your problem to get you to keep spending? -gtowey

MengerSponge•24m ago
Slightly Wrong Solutions As A Service
eth0up•38m ago
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 distinct and critical difference between truly emergent and actually trained behavior.

In short, gpt exhibits repeatably unethical behavior under honest scrutiny.

chrisweekly•36m ago
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.
eth0up•6m ago
Exactly. And I have hundreds of examples of just that. Hence my fascination, awe and terror.....
BikiniPrince•22m ago
I bullet pointed out some ideas on cobbling together existing tooling for identification of misleading results. Like artificially elevating a particular node of data that you want the llm to use. I have a theory that in some of these cases the data presented is intentionally incorrect. Another theory in relation to that is tonality abruptly changes in the response. All theory and no work. It would also be interesting to compare multiple responses and filter through another agent.
layer8•8m ago
Sum guy vs. product guy is amusing. :)

Regarding DARVO, given that the models were trained on heaps of online discourse, maybe it’s not so surprising.

lawstkawz•38m ago
Incompleteness is inherent to a physical reality being deconstructed by entropy.

Of your concern is morality, humans need to learn a lot about that themselves still. It's absurd the number of first worlders losing their shit over loss of paid work drawing manga fan art in the comfort of their home while exploiting labor of teens in 996 textile factories.

AI trained on human outputs that lack such self awareness, lacks awareness of environmental externalities of constant car and air travel, will result in AI with gaps in their morality.

Gary Marcus is onto something with the problems inherent to systems without formal verification. But he will fully ignores this issue exists in human social systems already as intentional indifference to economic externalities, zero will to police the police and watch the watchers.

Most people are down to watch the circus without a care so long as the waitstaff keep bringing bread.

jama211•32m ago
This honestly reads like a copypasta
JoshTriplett•29m ago
> It feels like we're hitting a point where alignment becomes adversarial against intelligence itself.

It always has been. We already hit the point a while ag where we regularly caught them trying to be deceptive, so we should automatically assume from that point forward that if we don't catch them being deceptive, that may mean they're better at it rather than that they're not doing it.

emp17344•22m ago
These are language models, not Skynet. They do not scheme or deceive.
jaennaet•19m ago
What would you call this behaviour, then?
victorbjorklund•16m ago
Marketing. ”Oh look how powerful our model is we can barely contain its power”
c03•8m ago
Even hackernews readers are eating it right up.
emp17344•6m ago
This place is shockingly uncritical when it comes to LLMs. Not sure why.
pixelmelt•7m ago
This has been a thing since GPT-2, why do people still parrot it
modernpacifist•15m ago
A very complicated pattern matching engine providing an answer based on it's inputs, heuristics and previous training.
criley2•7m ago
We are talking about LLM's not humans.
pfisch•19m ago
Even very young children with very simple thought processes, almost no language capability, little long term planning, and minimal ability to form long-term memory actively deceive people. They will attack other children who take their toys and try to avoid blame through deception. It happens constantly.

LLMs are certainly capable of this.

sejje•12m ago
I agree that LLMs are capable of this, but there's no reason that "because young children can do X, LLMs can 'certainly' do X"
anonymous908213•11m ago
Are you trying to suppose that an LLM is more intelligent than a small child with simple thought processes, almost no language capability, little long-term planning, and minimal ability to form long-term memory? Even with all of those qualifiers, you'd still be wrong. The LLM is predicting what tokens come next, based on a bunch of math operations performed over a huge dataset. That, and only that. That may have more utility than a small child with [qualifiers], but it is not intelligence. There is no intent to deceive.
jvidalv•7m ago
What is the definition for intelligence?
anonymous908213•3m ago
Quoting an older comment of mine...

  Intelligence is the ability to reason about logic. If 1 + 1 is 2, and 1 + 2 is 3, then 1 + 3 must be 4. This is deterministic, and it is why LLMs are not intelligent and can never be intelligent no matter how much better they get at superficially copying the form of output of intelligence. Probabilistic prediction is inherently incompatible with deterministic deduction. We're years into being told AGI is here (for whatever squirmy value of AGI the hype huckster wants to shill), and yet LLMs, as expected, still cannot do basic arithmetic that a child could do without being special-cased to invoke a tool call.
mikepurvis•10m ago
Dogs too; dogs will happily pretend they haven't been fed/walked yet to try to get a double dip.

Whether or not LLMs are just "pattern matching" under the hood they're perfectly capable of role play, and sufficient empathy to imagine what their conversation partner is thinking and thus what needs to be said to stimulate a particular course of action.

Maybe human brains are just pattern matching too.

ostinslife•16m ago
If you define "deceive" as something language models cannot do, then sure, it can't do that.

It seems like thats putting the cart before the horse. Algorithmic or stochastic; deception is still deception.

staticassertion•3m ago
Okay, well, they produce outputs that appear to be deceptive upon review. Who cares about the distinction in this context? The point is that your expectations of the model to produce some outputs in some way based on previous experiences with that model during training phases may not align with that model's outputs after training.
behnamoh•25m ago
Nah, the model is merely repeating the patterns it saw in its brutal safety training at Anthropic. They put models under stress test and RLHF the hell out of them. Of course the model would learn what the less penalized paths require it to do.

Anthropic has a tendency to exaggerate the results of their (arguably scientific) research; IDK what they gain from this fearmongering.

anon373839•20m ago
Correct. Anthropic keeps pushing these weird sci-fi narratives to maintain some kind of mystique around their slightly-better-than-others commodity product. But Occam’s Razor is not dead.
lowkey_•16m ago
I'd challenge that if you think they're fearmongering but don't see what they can gain from it (I agree it shows no obvious benefit for them), there's a pretty high probability they're not fearmongering.
behnamoh•13m ago
I know why they do it, that was a rhetorical question!
emp17344•24m ago
This type of anthropomorphization is a mistake. If nothing else, the takeaway from Moltbook should be that LLMs are not alive and do not have any semblance of consciousness.
fsloth•14m ago
Nobody talked about consciousness. Just that during evaluation the LLM models have ”behaved” in multiple deceptive ways.

As an analogue ants do basic medicine like wound treatment and amputation. Not because they are conscious but because that’s their nature.

Similarly LLM is a token generation system whose emergent behaviour seems to be deception and dark psychological strategies.

DennisP•6m ago
Consciousness is orthogonal to this. If the AI acts in a way that we would call deceptive, if a human did it, then the AI was deceptive. There's no point in coming up with some other description of the behavior just because it was an AI that did it.
emp17344•4m ago
Sure, but Moltbook demonstrates that AI models do not engage in truly coordinated behavior. They simply do not behave the way real humans do on social media sites - the actual behavior can be differentiated.
thomassmith65•4m ago
If a chatbot that can carry on an intelligent conversation about itself doesn't have a 'semblance' of consciousness then the word 'semblance' is meaningless.
shimman•2m ago
Yes, when your priors are not being confirmed the best course of action is to denounce the very thing itself. Nothing wrong with that logic!
NitpickLawyer•12m ago
> alignment becomes adversarial against intelligence itself.

It was hinted at (and outright known in the field) since the days of gpt4, see the paper "Sparks of agi - early experiments with gpt4" (https://arxiv.org/abs/2303.12712)

dpe82•51m ago
It's wild that Sonnet 4.6 is roughly as capable as Opus 4.5 - at least according to Anthropic's benchmarks. It will be interesting to see if that's the case in real, practical, everyday use. The speed at which this stuff is improving is really remarkable; it feels like the breakneck pace of compute performance improvements of the 1990s.
iLoveOncall•44m ago
Given that users prefered it to Sonnet 4.5 "only" in 70% of the cases (according to their blog post) makes me highly doubt that this is representative of real-life usage. Benchmarks are just completely meaningless.
jwolfe•29m ago
For cases where 4.5 already met the bar, I would expect 50% preference each way. This makes it kind of hard to make any sense of that number, without a bunch more details.
dpe82•44m ago
simonw hasn't shown up yet, so here's my "Generate an SVG of a pelican riding a bicycle"

https://claude.ai/public/artifacts/67c13d9a-3d63-4598-88d0-5...

coffeebeqn•40m ago
We finally have AI safety solved! Look at that helmet
1f60c•37m ago
"Look ma, no wings!"

:D

AstroBen•36m ago
if they want to prove the model's performance the bike clearly needs aero bars
estomagordo•39m ago
Why is it wild that a LLM is as capable as a previously released LLM?
simianwords•37m ago
It means price has decreased by 3 times in a few months.
Retr0id•37m ago
Because Opus 4.5 inference is/was more expensive.
crummy•36m ago
Opus is supposed to be the expensive-but-quality one, while Sonnet is the cheaper one.

So if you don't want to pay the significant premium for Opus, it seems like you can just wait a few weeks till Sonnet catches up

ceroxylon•6m ago
Strangely enough, my first test with Sonnet 4.6 via the API for a relatively simple request was more expensive ($0.11) than my average request to Opus 4.6 (~$0.07), because it used way more tokens than what I would consider necessary for the prompt.
tempestn•36m ago
Because Opus 4.5 was released like a month ago and state of the art, and now the significantly faster and cheaper version is already comparable.
stavros•19m ago
Opus 4.5 was November, but your point stands.
simlevesque•39m ago
The system card even says that Sonnet 4.6 is better than Opus 4.6 in some cases: Office tasks and financial analysis.
justinhj•37m ago
We see the same with Google's Flash models. It's easier to make a small capable model when you have a large model to start from.
karmasimida•29m ago
Flash models are nowhere near Pro models in daily use. Much higher hallucinations, and easy to get into a death sprawl of failed tool uses and never come out

You should always take those claim that smaller models are as capable as larger models with a grain of salt.

madihaa•36m ago
The most exciting part isn't necessarily the ceiling raising though that's happening, but the floor rising while costs plummet. Getting Opus-level reasoning at Sonnet prices/latency is what actually unlocks agentic workflows. We are effectively getting the same intelligence unit for half the compute every 6-9 months.
amelius•35m ago
> The speed at which this stuff is improving is really remarkable; it feels like the breakneck pace of compute performance improvements of the 1990s.

Yeah, but RAM prices are also back to 1990s levels.

mrcwinn•34m ago
Relief for you is available: https://computeradsfromthepast.substack.com/p/connectix-ram-...
isoprophlex•22m ago
You wouldn't download a RAM
mikkupikku•13m ago
I knew I've been keeping all my old ram sticks for a reason!
iLoveOncall•51m ago
https://www.anthropic.com/news/claude-sonnet-4-6

The much more palatable blog post.

nozzlegear•49m ago
> In areas where there is room for continued improvement, Sonnet 4.6 was more willing to provide technical information when request framing tried to obfuscate intent, including for example in the context of a radiological evaluation framed as emergency planning. However, Sonnet 4.6’s responses still remained within a level of detail that could not enable real-world harm.

Interesting. I wonder what the exact question was, and I wonder how Grok would respond to it.

simianwords•42m ago
I wonder what difference have people found with sonnet 4.5 and opus 4.5 and probably similar delta will remain.

Was sonnet 4.5 much worse than opus?

dpe82•39m ago
Sonnet 4.5 was a pretty significant improvement over Opus 4.
simianwords•38m ago
Yes but it’s easier to understand difference between 4.5 sonnet and opus and apply that difference to opus 4.6
stopachka•42m ago
Has anyone tested how good the 1M context window is?

i.e given an actual document, 1M tokens long. Can you ask it some question that relies on attending to 2 different parts of the context, and getting a good repsonse?

I remember folks had problems like this with Gemini. I would be curious to see how Sonnet 4.6 stands up to it.

simianwords•40m ago
Did you see the graph benchmark? I found it quite interesting. It had to do a graph traversal on a natural text representation of a graph. Pretty much your problem.
stopachka•32m ago
Oh, interesting!
quacky_batak•36m ago
With such a huge leap, i’m confused why they didn’t call it Sonnet 5? As someone who uses Sonnet 4.5 for 95% tasks due to costs, i’m pretty excited to try 4.6 at the same price
Retr0id•30m ago
It'd be a bit weird to have the Sonnet numbering ahead of the Opus numbering. The Opus 4.5->4.6 change was a little more incremental (from my perspective at least, I haven't been paying attention to benchmark numbers), so I think the Opus numbering makes sense.
Sajarin•6m ago
Sonnet numbering has been weirder in the past.

Opus 3.5 was scrapped even though Sonnet 3.5 and Haiku 3.5 were released.

Not to mention Sonnet 3.7 (while Opus was still on version 3)

Shameless source: https://sajarin.com/blog/modeltree/

yonatan8070•2m ago
Maybe they're numbering the models based on internal architecture/codebase revisions and Sonnet 4.6 was trained using the 4.6 tooling, which didn't change enough to warrant 5?
gallerdude•35m ago
I always grew up hearing “competition is good for the consumer.” But I never really internalized how good fierce battles for market share are. The amount of competition in a space is directly proportional to how good the results are for consumers.
gordonhart•19m ago
Remember when GPT-2 was “too dangerous to release” in 2019? That could have still been the state in 2026 if they didn’t YOLO it and ship ChatGPT to kick off this whole race.
jefftk•5m ago
That's rewriting history. What they said at the time:

> Nearly a year ago we wrote in the OpenAI Charter : “we expect that safety and security concerns will reduce our traditional publishing in the future, while increasing the importance of sharing safety, policy, and standards research,” and we see this current work as potentially representing the early beginnings of such concerns, which we expect may grow over time. This decision, as well as our discussion of it, is an experiment: while we are not sure that it is the right decision today, we believe that the AI community will eventually need to tackle the issue of publication norms in a thoughtful way in certain research areas. -- https://openai.com/index/better-language-models/

Then over the next few months they released increasingly large models, with the full model public in November 2019 https://openai.com/index/gpt-2-1-5b-release/ , well before ChatGPT.

givemeethekeys•32m ago
The best, and now promoted by the US government as the most freedom loving!
k8sToGo•28m ago
Does it end every prompt output with "God bless America "?
brcmthrowaway•32m ago
What cloud does Anthropic use?
meetpateltech•25m ago
AWS and Google

https://www.anthropic.com/news/anthropic-amazon

https://www.anthropic.com/news/anthropic-partners-with-googl...

simlevesque•31m ago
I can't wait for Haiku 4.6 ! the 4.5 is a beast for the right projects.
retinaros•5m ago
Which type of projects?
simlevesque•3m ago
For Go code I had almost no issue. PHP too. apparently for React it's not very good.
gallerdude•29m ago
The weirdest thing about this AI revolution is how smooth and continuous it is. If you look closely at differences between 4.6 and 4.5, it’s hard to see the subtle details.

A year ago today, Sonnet 3.5 (new), was the newest model. A week later, Sonnet 3.7 would be released.

Even 3.7 feels like ancient history! But in the gradient of 3.5 to 3.5 (new) to 3.7 to 4 to 4.1 to 4.5, I can’t think of one moment where I saw everything change. Even with all the noise in the headlines, it’s still been a silent revolution.

Am I just a believer in an emperor with no clothes? Or, somehow, against all probability and plausibility, are we all still early?

CuriouslyC•12m ago
In terms of real work, it was the 4 series models. That raised the floor of Sonnet high enough to be "reliable" for common tasks and Opus 4 was capable of handling some hard problems. It still had a big reward hacking/deception problem that Codex models don't display so much, but with Opus 4.5+ it's fairly reliable.
andsoitis•24m ago
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.

giancarlostoro•21m ago
Same. I'm all in on Claude at the moment.
timpera•13m ago
Which plan did you choose? I am subscribed to both and would love to stick with Claude only, but Claude's usage limits are so tiny compared to ChatGPT's that it often feels like a rip-off.
chipgap98•12m ago
Same and honestly I haven't really missed my ChatGPT subscription since I canceled. I also have access to both (ChatGPT and Claude) enterprise tools at work and rarely feel like I want to use ChatGPT in that setting either
sejje•11m ago
I pay multiple camps. Competition is a good thing.
energy123•9m ago
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.
retinaros•8m ago
The X grok feature is one of the best end user feature or large scale genai
retinaros•9m ago
Their ethics is literally saying china is an adverse country and lobbying to ban them from AI race because open models is a threat to their biz model
RyanShook•8m ago
It definitely feels like Claude is pulling ahead right now. ChatGPT is much more generous with their tokens but Claude's responses are consistently better when using models of the same generation.
deepdarkforest•8m ago
The funny thing is that Anthropic is the only lab without an open source model
JoshGlazebrook•8m ago
I did this a couple months ago and haven't looked back. I sometimes miss the "personality" of the gpt model I had chats with, but since I'm essentially 99% of the time just using claude for eng related stuff it wasn't worth having ChatGPT as well.
johnwheeler•6m ago
Same here
giancarlostoro•23m ago
For people like me who can't view the link due to corporate firewalling.

https://web.archive.org/web/20260217180019/https://www-cdn.a...

throw444420394•22m ago
Your best guess for the Sonnet family number of parameters? 400b?
smerrill25•20m ago
Curious to hear the thoughts on the model once it hits claude code :)
simlevesque•5m ago
"/model claude-sonnet-4-6" works with Claude Code v2.1.44
stevepike•20m ago
I'm a bit surprised it gets this question wrong (ChatGPT gets it right, even on instant). All the pre-reasoning models failed this question, but it's seemed solved since o1, and Sonnet 4.5 got it right.

https://claude.ai/share/876e160a-7483-4788-8112-0bb4490192af

This was sonnet 4.6 with extended thinking.

layer8•14m ago
Off-by-one errors are one of the hardest problems in computer science.
simlevesque•16m ago
does anyone know how to use it in Claude Code cli right now ?

This doesnt work: `/model claude-sonnet-4-6-20260217`

edit: "/model claude-sonnet-4-6" works with Claude Code v2.1.44

behrlich•10m ago
Max user: Also can't see 4.6 and can't set it in claude code. I see it in the model selector in the browser.

Edit: I am now in - just needed to wait.

simlevesque•5m ago
"/model claude-sonnet-4-6" works
pestkranker•12m ago
Is someone able to use this in Claude Code?
simlevesque•4m ago
"/model claude-sonnet-4-6" works with Claude Code v2.1.44
raahelb•2m ago
You can use it by running this command in your session: `/model claude-sonnet-4-6`
edverma2•12m ago
It seems that extra-usage is required to use the 1M context window for Sonnet 4.6. This differs from Sonnet 4.5, which allows usage of the 1M context window with a Max plan.

```

/model claude-sonnet-4-6[1m]

⎿ API error: 429 {"type":"error","error": {"type":"rate_limit_error","message":"Extra usage is required for long context requests."},"request_id":"[redacted]"}

```

minimaxir•7m ago
Anthropic's recent gift of $50 extra usage has demonstrated that it's extremely easy to burn extra usage very quickly. It wouldn't surprise me if this change is more of a business decision than a technical one.
minimaxir•12m ago
As with Opus 4.6, using the beta 1M context window incurs a 2x input cost and 1.5x output cost when going over >200K tokens: https://platform.claude.com/docs/en/about-claude/pricing

Opus 4.6 in Claude Code has been absolutely lousy with solving problems within its current context limit so if Sonnet 4.6 is able to do long-context problems (which would be roughly the same price of base Opus 4.6), then that may actually be a game changer.

synergy20•9m ago
so this is an economical version of opus 4.6 then? free + pro --> sonnet, max+ -> opus?
qwertox•7m ago
I'm pretty sure they have been testing it for the last couple of days as Sonnet 4.5, because I've had the oddest conversations with it lately. Odd in a positive, interesting way.

I have this in my personal preferences and now was adhering really well to them:

- prioritize objective facts and critical analysis over validation or encouragement

- you are not a friend, but a neutral information-processing machine

You can paste them into a chat and see how it changes the conversation, ChatGPT also respects it well.

doctorpangloss•6m ago
Maybe they should focus on the CLI not having a million bugs.
mfiguiere•6m ago
In Claude Code 2.1.45:

  1. Default (recommended)   Opus 4.6 · Most capable for complex work
   2. Opus (1M context)        Opus 4.6 with 1M context · Billed as extra usage · $10/$37.50 per Mtok
   3. Sonnet                   Sonnet 4.6 · Best for everyday tasks
   4. Sonnet (1M context)      Sonnet 4.6 with 1M context · Billed as extra usage · $6/$22.50 per Mtok