Understand this story
On November 4, 2026, UCSD's Hao AI Lab announced that it would integrate Pika Labs' open-source MiniMax H3 Flex into FastVideo. According to its published information, the official open-source inference code takes 59.7 seconds to generate a 5-second 480p video with audio on a single GB200, while FastVideo BF16 takes 26.6 seconds, NVFP4 QAD takes 16.8 seconds, and FP4 attention takes 15.2 seconds. Hao AI Lab claims it is up to 7 times faster than the official code, and added multi-GPU support, with 4 GPUs generating a 15-second clip 7 times faster. NVFP4 QAD distills BF16 Flex, coming 14-17% closer to BF16 than ordinary quantization; FP4 attention reduces a 768p 15-second clip from 101.8 seconds to 85.3 seconds. That clip takes 59.5 seconds on a single DGX Spark. FastVideo also supports the Open-H3-Regenerate-2K 2K model, and the weights, inference code, and complete training recipe have all been released.
AI-generated from public reports
Latest development · Report
Hao AI Lab integrates MiniMax H3 Flex into FastVideo, single...Developments
1 confirmed developments
- Source:UCSD Hao AI Lab
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 4
UCSD Hao AI LabFirst-partySelected
Hao AI Lab integrates MiniMax H3 Flex into FastVideo, single...
Hao AI Lab has integrated Pika Labs' open-source MiniMax H3 Flex into FastVideo. The official open-source inference code takes 59.7 seconds to generate a 5-second 480p video with audio on a single GB200, while FastVideo BF16 takes 26.6 seconds, NVFP4 QAD takes 16.8 seconds, and FP4 attention takes 15.2 seconds. The team claims it is up to 7 times faster than the official code, and has added multi-GPU support, generating a 15-second clip 7 times faster on 4 GPUs. NVFP4 QAD distills the BF16 Flex model, coming 14-17% closer to BF16 than ordinary quantization; FP4 attention reduces a 768p 15-second clip from 101.8 seconds to 85.3 seconds. This clip takes 59.5 seconds on a single DGX Spark. FastVideo also supports the 2K model of Open-H3-Regenerate-2K, and the weights, inference code, and complete training recipe have all been released.