Rodrigo Ornellas

Developer, AI Enginner (in love with AI), thinker

[Rodrigo Ornellas @RodrigOrnellas]

@theo luna and flash above grok 4.6?
well, your experience with these models are def very different of mine.
luna only does crazy stuff in my projects.

[Rodrigo Ornellas @RodrigOrnellas]

@Da7_Tech that means it needs a lot

[Rodrigo Ornellas @RodrigOrnellas]

@Ashleyo4h of course not!
who would be stupid to take sucha crazy deal?

[Rodrigo Ornellas @RodrigOrnellas]

@eunaativa @grok , isso existe? faz mal para a pele?

[Rodrigo Ornellas @RodrigOrnellas]

@CampbellKaleb23 @thsottiaux not yet as well... if so many ppl didn't get it yet, it might mean they did not released it yet

[Rodrigo Ornellas @RodrigOrnellas]

@thsottiaux not sure if the banked resets were given already, but I did not get it.
can you tell about it? is it yet to happen or I just did not get my banked reset?

[Rodrigo Ornellas @RodrigOrnellas]

@Ananth7e yeah... I noticed it too.. haha
just weird, but he can still have days left for his subs

[Rodrigo Ornellas @RodrigOrnellas]

@theo I did not get any bank reset.

[Rodrigo Ornellas @RodrigOrnellas]

@_thomasip @blueemi99 yeah... there is this one:

https://t.co/kYFHYttuH0

he is the person who wrote deep swe paper and he said he ran in a bigger subset and got 63%

[Rodrigo Ornellas @RodrigOrnellas]

@theo I am very curious to hear about your internal benchmarks Theo

[Rodrigo Ornellas @RodrigOrnellas]

@elonmusk @yishan next step is to put it into the sun

[Rodrigo Ornellas @RodrigOrnellas]

@OpenRouter @poteto said that yesterday would be exciting.. perhaps its some cursor related model such as composer or grok

[Rodrigo Ornellas @RodrigOrnellas]

@MiaAI_lab blink if I am right: is it composer-3?

[Rodrigo Ornellas @RodrigOrnellas]

@ryanvogel I know it seems to be GLM... but it could be composer-3, which is a model everyone is waiting for.

[Rodrigo Ornellas @RodrigOrnellas]

@MiaAI_lab if it's not GLM, according to what I have seen, so someone copied it.

Comparison modelExact matches with Ox Alpha
GLM-5.260/60
Kimi K2.68/60
Hy32/60
Qwen 3.52/60
DeepSeek V41/60
MiniMax M31/60
MiMo 2.50/60

[Rodrigo Ornellas @RodrigOrnellas]

@Ananth7e Ananth, what the "x" mean? zero score or "not 4/4"?

[Rodrigo Ornellas @RodrigOrnellas]

@GrokInsider As an old user, I can tell you that cursor has evolved significantly and it provides one of the best user experiences atm. Def worth your time.

[Rodrigo Ornellas @RodrigOrnellas]

@Cyibarumawe what the hell is it @grok ?

[Rodrigo Ornellas @RodrigOrnellas]

@fahnx8 there is a thing called market. there is other thing called value.
doesn't matter the price. market must to see value in it, it's not well priced then.
that's why ppl complaint. it's not about the gross api token value. it's the value it does deliver.