FastVideo System Requirements (CPU, RAM, Disk)

FastVideo is a unified inference and post-training framework from Hao AI Lab for accelerated video generation. It supports high-performance inference for popular video diffusion models, plus a full stack of post-training recipes including SFT, LoRA, DPO and distillation. With FlashAttention, Tensor Parallel and advanced sampling, FastVideo delivers fast generation and efficient fine-tuning on a single GPU or a whole cluster — for text-to-video and video editing alike.