FastFrames video generation model released
The story so far
Hugging Face released FastFrames, a set of fast autoregressive video diffusion models, achieving interactive latency at a speed of 36ms/frame (28RPS) on a single B200 GPU while matching the generation quality of nearly 8 times slower models. Its 14B version runs at 45ms/frame (22RPS) on an 8x B200 GPU node, significantly improving visual quality and realism.
AI-written from the reports · Updated 11 hours ago
Coverage timeline
Follow the coverage to see every side of the story.
- Hugging Face Daily PapersFastFrames: Video Generation at Interactive Latency
FastFrames is a set of fast autoregressive video diffusion models, achieving interactive latency at a speed of 36ms/frame (28RPS) on a single B200 GPU while matching the generation quality of nearly 8 times slower models; its 14B version runs at 45ms/frame (22RPS) on an 8x B200 GPU node, significantly improving visual quality and realism.
Heat over time
- Now
- 6
- Peak
- 718 Nov, 19:00
- 24h change
- –
Hover or tap the chart to see each hour; use the arrow keys on a keyboard.