Ultra Flash: Scaling Real-Time Streaming Video Generation to High Resolutions
2026-06-08 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors introduce Ultra Flash, a new method that can generate high-resolution videos in real-time using just one GPU. They improve on previous models by adding a special super-resolution training step, a clever way to upscale video details while keeping the video smooth, and an optimization process that makes the video generation faster and more coherent. Their experiments show Ultra Flash can produce detailed videos quickly, at resolutions much higher than before. This helps move real-time high-quality video generation closer to practical use.
autoregressive video diffusionsuper-resolutionreal-time video generationlatent upsamplerhigh-resolution decodercausal streamingsingle GPUspatiotemporal coherencemodel distillationself-forcing optimization
Authors
Luxury, Jie Huang, Zihao Fan, Xiaoxiao Ma, Yuming Li, Jun-hao Zhuang, Zeyue Xue, Siming Fu, Haoran Li, Mingchen Zhong, Guohui Zhang, Shichen Ma, Yijun Liu, Jiaqi Shi, Yanwen Ma, Yaofeng Su, Haoyu Wang, Yaowei Li, Songchun Zhang, Weiyang Jin, Yuxuan Bian, Shiyi Zhang, Haojun Xu, Shuai Lu, Xin Han, Wei Tang, Haoyang Huang, Nan Duan
Abstract
While recent autoregressive video diffusion models achieve remarkable streaming quality, they remain confined to low resolutions (e.g., 480P), leaving efficient, scalable, real-time high-resolution video generation a fundamental open challenge. To bridge this gap, we present Ultra Flash, a cascaded streaming framework capable of real-time high-resolution video generation. Ultra Flash achieves ~30 FPS at 1K resolution and ~18 FPS at 2K resolution on a single GPU through three key contributions: (1) an architecture-preserving T2V-to-TV2V super-resolution training paradigm coupled with an AIGC-oriented data degradation pipeline that effectively preserves the generative capability of the base model, enabling enhanced high-resolution detail when cascaded after mainstream low-resolution generative models; (2) a causal streaming latent upsampler paired with a high-resolution decoder, which enhances spatiotemporal coherence while enabling efficient latent spatial scaling and precise high-resolution decoding with negligible computational overhead; and (3) a cascade high-resolution streaming video generation optimization scheme that first performs hybrid-reward-enhanced sparse causalization and single-step distillation of the super-resolution model, then introduces cascaded streaming self-forcing preference optimization with dynamic cache management, jointly enhancing overall coherence, improving quality, and enabling real-time high-resolution streaming video generation. Extensive experiments demonstrate that Ultra Flash reliably produces ultra-high-resolution streaming video while maintaining state-of-the-art visual quality and superior efficiency.