Clarifai achieves 414 tokens per second on Kimi K2.5, one of the first providers to reach 400+ TPS on a trillion-parameter reasoning model running on Nvidia B200 GPUs.
Clarifai achieves 414 tokens per second on Kimi K2.5, one of the first providers to reach 400+ TPS on a trillion-parameter reasoning model running on Nvidia B200 GPUs.