3000 tokens/sec LLM playground
playground.kog.aiActually getting to use this is different than just reading about it. Feels like a preview of the future.
Related discussion of their writeup: https://news.ycombinator.com/item?id=48321076
What am I missing? It can't even list countries in NATO correctly, chatjimmy.ai can do that and output 17000 tokens/sec on cheaper silicon.
I don't think you're missing anything, the speed is the impressive thing but the model is limited. Perhaps Chatjimmy is more impressive in this respect. Most people haven't tried either