Remix.run Logo
dools 2 hours ago

Kimi k2.6 is about on par with GPT 5.2 so I’d say open weight models are about 6 months behind.

cbg0 2 hours ago | parent | next [-]

The Q4 quantization requires about 600GB of RAM without context, not exactly consumer hardware friendly.

janderland 2 hours ago | parent | prev [-]

Has Kimi found a way to vastly reduce the amount of VRAM required without running at 3 tokens per second? That’s the real concern.