@emanueledpt Sonnet is still absolute beast of a model, with so much usage in sub plans. If it comes even close to Opus 4.8, it'll be an amazing model for loops and goals tasks.
@notjazii Its also not confirmed if 5.6 also gonna be worldwide though, Altman said working hard for it, doesn't mean its confirmed...
In pure performace Fable looks still stronger, but in availability? 5.6 might take it if it comes to all and without usage credit.
@ZixuanLi_ As you guys used to focus on that hence we got great long form writing models like 4.7 but i guess the priority got steered away. Bring the priority back to creative writing too.
@ZixuanLi_ Actually good in creative Writing and Roleplaying. GLM has been decent at it but insane positive bias and force softening everything has been ruining the experience. Please improve it as there is a *massive* market for that too! That you guys can fully capture.
@HarshithLucky3 Gemini models are *the* best in knowledge type of works. If you wanna learn something or just ask about random things, myth busting in general. And its vision is really good too.
@haider1 bruh, Mythos is *not* for us at all. They are for companies, Fable is the one for general use. Its the same model with more guardrails. They literally said that in the blogpost lmfao , π€¦, X is just full of retards.