SANA-Video_2.0_5B_720p_4step

Research preview: This distilled checkpoint is an early T2V-only preview. For the original 50-step T2V + TI2V release, use SANA-Video_2.0_5B_720p.

Project page · Online demo · Documentation · GitHub · Paper

SANA-Video 2.0 is an efficient diffusion transformer for high-resolution video generation. This preview provides a full-model EMA checkpoint distilled for four-stage text-to-video generation at 720p. It combines gated bidirectional linear-attention layers with periodic dense softmax-attention anchors and shared Attention Residual aggregation.

Model details

Checkpoint lineage and format

The selected checkpoint is the global-step-1000 DMD EMA export initialized from the SANA-Video 2.0 SFT model after merging the ReFL step-500 adapter. DMD then updates the full transformer; this release is therefore a full model, not a LoRA adapter.

The checkpoint contains a state_dict_ema tensor mapping only. It does not contain optimizer, learning-rate scheduler, gradient-scaler, or training-loop state. The inference entry point unwraps state_dict_ema, removes an optional model. key prefix, and casts the transformer to BF16.

Files

• checkpoints/SANA_Video_2.0_5B_720p_4step.pth: distilled EMA transformer
• config.yaml: clean 5B source-tower config with the preview frame/FPS defaults
• demo/: verified seed-4 MP4 and poster generated from this checkpoint
• LICENSE: Apache License 2.0

Verified 4-step example

This 1280 × 736, 81-frame, 16 FPS sample was generated from the released checkpoint with seed 4 and the exact command shown below.

Prompt: In a cozy, vintage room adorned with floral wallpaper, a cartoon rooster sits comfortably in a floral-patterned armchair, sipping from a bottle of beer. The rooster, with its vibrant red comb and wattle, displays a range of expressions—smiling, nodding, and opening its beak wide in a cheerful manner. The setting includes wooden furniture and another beer bottle on the table, adding to the relaxed atmosphere. The camera captures the rooster from a close-up angle, emphasizing its animated movements and lively demeanor.

Four-stage sampling contract

This model does not use a truncated DPM-Solver trajectory. At each fixed stage, the model predicts velocity v, computes x0 = x - sigma * v, and—except at the terminal stage—re-noises x0 at the next sigma with a fresh draw from the same seeded generator.

The rollout uses one BF16 initial-noise draw and three BF16 transition-noise draws. flow_shift is not applied by this sampler.

Inference

Until the upstream PR is merged, use the preview branch:

git clone https://github.com/Efficient-Large-Model/Sana.git
cd Sana
git checkout feat/sana-video2-4step-preview
bash environment_setup.sh sana
conda activate sana

Place the Diffusers-format LTX 2.3 VAE at output/pretrained_models/LTX-2.3-Diffusers/, or update vae.vae_pretrained in the config.

The following command uses the exact public preview settings:

bash inference_video_scripts/inference_sana_video.sh \
  --np 1 \
  --config configs/sana_video2/SanaVideo2_5B_720p.yaml \
  --model_path hf://Efficient-Large-Model/SANA-Video_2.0_5B_720p_4step/checkpoints/SANA_Video_2.0_5B_720p_4step.pth \
  --txt_file=asset/samples/sana_video2_5b_720p_demo.txt \
  --task=t2v \
  --model.image_size=480 \
  --custom_height_width='[736,1280]' \
  --sampling_algo=fastvideo_dmd_4step \
  --generator_sigma_profile=sana_shift6_dpm \
  --cfg_scale=1.0 \
  --flow_shift=1.0 \
  --motion_score=0 \
  --negative_prompt=None \
  --num_frames=81 \
  --step=4 \
  --fps=16 \
  --seed=4 \
  --work_dir output/sana_video2_t2v_720p_4step_preview

The selected global-step-1000 checkpoint was evaluated on the full VBench T2V suite: 4,730 prompts across all 16 dimensions, using the same four-stage sampling contract, 736×1280 resolution, 81 frames, 16 FPS, and CFG 1.

Intended use

This model is intended for research, evaluation, and creative generation of short videos from text. It can also serve as a full-model starting point for further research under the license terms. It is not intended for first-frame image conditioning; use the base release for TI2V.

The model is not intended to produce factual evidence, identify people, make high-impact automated decisions, or generate content that violates privacy, copyright, applicable law, or platform policies.

Limitations and bias

• This is a research preview and has less extensive release validation than the original 50-step model.
• Generated motion, anatomy, text rendering, object permanence, and physical interactions may be inconsistent, especially for crowded or highly dynamic s