@0x0SojalSec@TechMDAI People wanna model that can do what they need quickly enough and the fact of the matter is that most open weight models now can do that for much cheaper while offering the user more control and stability when hosted locally
@varrock Pro tip whenever youโre paying a lot of money for the same thing that can be had for cheaper elsewhere youโre mostly paying for exclusivity and curation of other participants.
@TechMDAI DSV4.1F achieves higher decode and Prefill speeds with bigger context and multimodal. Intelligence takes a little more thinking tokens to lead to a good outcome than GLM 5.3 but itโs worth it because most tasks donโt need that much thought.
@eurofounder my favorite part of Europe is where they run AC eco mode with the doors and windows wide open all day then preach about saving the environment