if you're making a game with ai npcs, the worst part is standing there waiting for them to talk

just like I did with Bully, you can link the npcs in your games up to an ai model running on a rented nvidia b300 with eagle-4 on top

eagle-4 is basically a tiny model that guesses the next few words, then the big model checks them all at once instead of writing one word at a time

same answers, just faster

the voice and game side don't get faster but just the part where the npc is thinking

if you want to run those models on actual physical NVIDIA server GPUs, you can run that and then have it plugged into whatever you're doing