AI Paper Atlas

Research map / Video / Video generation

SANA-Video: Efficient Video Generation with Block Linear Diffusion Transformer

Junsong Chen, Yuyang Zhao, Jincheng YU, Ruihang Chu, Junyu Chen, Shuai Yang, Xianbang Wang, Yicheng Pan, Daquan Zhou, Huan Ling, Haozhe Liu, Hongwei Yi and 8 more

ICLR 2026 · Video › Video generation · approach: Efficient video diffusion models · 84 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace full attention with linear attention (ablation)SANA-Video · 720p video generationlatency: 4x speedup at 720PImproved
replace full attention with linear attention (ablation)SANA-Video · 480p video generationlatency: 2x speedup at 480PImproved
add 3D RoPE after ReLU in linear attention (ablation)SANA-Video · training losstraining lossImproved
add 1D temporal convolution to Mix-FFN (ablation)SANA-Video · training loss and motiontraining lossImproved
use monotonically increasing SNR timestep sampler instead of random (ablation)LongSANA · autoregressive block trainingvideo quality and block consistencyImproved
remove RoPE from key or query in the denominator of linear attention (ablation)SANA-Video · training stabilitytraining stabilityImproved
use DCAE-V instead of Wan2.1VAE under latent noise 0.2VAE reconstruction · Panda-70M 1000 samplesPSNR: 24.25 -> 29.34Improved
use DCAE-V instead of Wan2.2VAE under latent noise 0.2VAE reconstruction · Panda-70M 1000 samplesPSNR: 25.94 -> 29.34Improved
use DCAE-V instead of Wan2.2VAE without noiseVAE reconstruction · Panda-70M 192pPSNR: 35.61 -> 33.25Hurt
quantize SANA-Video to NVFP4 via SVDQuantSANA-Video · 720p 5s video on RTX 5090generation time: 71 s to 29 sImproved
swap Wan2.1 VAE for DCAE-V (720p)SANA-Video-2B · 720x1280x81 latencylatency: 36 vs 116 (Wan-2.2-5B)Improved

Later papers that compared with this method

4 later accepted papers report SANA-Video as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy