正在加载视频...
视频加载失败
sonnet 5 is routing to sonnet 5.5 it's been a week since it started routing to newer model and team already confirmed that sonnet 5.5 is coming soon from early testing it looks like a huge upgrade from sonnet 5 like opus 5 to opus 5.5 just look at... show more
57 条评论
is sonnet mogs sol, it's over for openai.
They won't let it
f Sonnet actually mogs Sol in real usage and Ant. drops it around DevDay, then OAI will have to accelerate 6.1 Astra and push an update to the Sol and Luna line like they did with the 5.6 series in august
they better drop better sol on dev day
When's the 10k post coming!!
tomorrow bro
But what about the use cases? In what context do you use it?
just small tests for now gonna have to proper testing soon but yeah
I abandoned Fable 5.1 completely, it’s super awesome!! I find Opus 5.5 to be much better even in terms of reasoning and logic, and I wonder… what will the next Fable be? A monster?
Yeah same for me too next fable would be 5.5 And it will be better then opus 5.5 bit expensive you know
Yes, but what interests me is a model’s intelligence in terms of investigation, intuition, finding and seeing the non-obvious. That’s where the revolution lies—not just a simple prediction, but changing the game entirely… How exciting the future is… Have you seen Meta’s announcements? I really think Mark is one of the few truly bringing innovation, yet at the same time, in my opinion, very misunderstood! I would really love to have a Muse Charm; the real revolution would be proactivity, with it seeking you out, like a truly living interaction.
No way!! Sonnet 5.5 is goal too haha
same reaction bro
6 sol better lol
Imagine if sonnet mogs Astra
If sonnet mogs Sol that’s it. I would be completely switching to Claude. I have 2 codex subscriptions, and Astra is too expensive for plus. If sonnet is this good, I would have no other logical choice.
Imagine Haiku 5.5 > 6 sol 🥲
nah that can't be true
I don't hope so lol
Sonnet seems to be better
you’re team openai? it’s not looking so good right now lol
i am team whoever gives me best model
At least for this, it is not. It's just different style.
im fkn jealous man. how's sonnet 5? would LOVE to hear more on ur experience working with it. and would be nice if you could do some measurements, re token efficiency and stuff like that
ask it do you know tibo the reset guy? Don't search in new chat
good point shouldve asked directly for codex reset guy or mention tibo. still not sadly but tysm man:)
I noticed Anthropic models are always the best in the fields that people publicly test them on. People do tests on games, voxels, 3d, SVGs and that's where it performs the best. I also feel like Astra mogs opus in logic and backend coding which lets me believe Claude 'hypemaxxxes' their models by training them on those specific fields.
that would be crazy 🤯
i am hearing anthropic wanna release this tomorrow
🔥 they're pacing the other way!
Today or tomorrow we may be getting it!!
nah maybe in like week or two
Sol's version makes a great landing page
Hmm lemme try
What....? It looks better than sol .. 😭
why it looks better?
Looks better to me bcz of the level of detail
Dairo is coming anthropic has been cooking all this while 😂
just imagine the anthropic developers
I am ant dev
Looking so cool
Which one?
Both are but I think sol good
Sonnet 5.5 sounds bangoooooor
Imagine being better than opus, as they did it before, lol
won't be better then opus but then sonnet 5
Well sonnet models surpassed opus ones before, but this might be too early
The routing rumors always make me refresh the playground lol. If 5.5 really closes the gap like that, it would be a nice surprise. Still hoping OpenAI pulls off a comeback too.
Considering Sol is Terra now and they api pricing is exactly the same, it doesn’t sound that bad, right? I mean Claude models suppose to be better, always been.
tbf 6 'sol' is more like the new terra tbf, I don't think that model deserves same recognition as 5.6 Sol. And the pricing is same as Sonnet 5, so if it beats 6 sol then it won't be too bad since Astra Minor is also coming.
It's ture, Sonnet now knows Tibo the reset guy without search
Only for some people
dm me I need a devin plan... 🫠
Bro gotta check DM once
Like all models bevor, im in the testing pool. I asked sonnet 5.5 the date of his training and he answered end of August 26. Opus 5.5 was june 26 and fable 5.x was July 26 @notjazii
Interesting man
did they just master post training and are fixing all their models?
