Total runs:
55.5K
Run Growth:
-59.7K
Growth Rate:
-107.64%
Sponsored by
Vmake Labs Video Enhancer
Efficient-Large-Model / SANA-Video_2.0_5B_720p_4step
Model's Last Updated: September 08 2026
text-to-video
Introduction of SANA-Video_2.0_5B_720p_4step
Model Details of SANA-Video_2.0_5B_720p_4step
SANA-Video 2.0 5B 720p — 4-Step Preview
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
| Property | Value |
|---|---|
| Architecture |
SanaVideo2_5B
|
| Parameters | 4,466,980,960 model parameters (4.47B) |
| Transformer | 32 layers, hidden size 2,560 |
| Attention | 75% gated linear attention, 25% dense softmax anchors |
| Attention Residuals | Shared, timestep-independent aggregation every 8 layers |
| Task | Text-to-video only |
| Output bucket | 736 × 1280, 81 frames, 16 FPS (about 5.06 seconds) |
| Model construction | Source tower at 480; explicit dynamic 736 × 1280 bucket |
| Text encoder |
google/gemma-2-2b-it
|
| VAE |
LTX 2.3, 128 latent channels, stride
(8, 32, 32)
|
| Inference |
BF16, CFG 1, 4 stages,
sana_shift6_dpm
, no motion suffix
|
| License | Apache 2.0 |
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.
▶ Watch or download the verified four-step video
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.
| Stage | Physical sigma | API timestep | Model embedding timestep |
|---|---|---|---|
| 1 | 0.9998332262 | 999.8332262 | 999 |
| 2 | 0.9471688271 | 947.1688271 | 947 |
| 3 | 0.8568977118 | 856.8977118 | 856 |
| 4 | 0.6663702130 | 666.3702130 | 666 |
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 scenes.
- Prompt following can degrade for long, ambiguous, or compositionally complex instructions.
- Outputs can reflect social and cultural biases present in training data and in the separately loaded text encoder.
- The model does not independently verify whether generated content is factual, safe, or free of third-party rights.
Users should review outputs before publication, disclose synthetic media where appropriate, and add safeguards suited to their application.
Resources
- SANA repository
- SANA-Video 2.0 four-step preview PR
- SANA-Video 2.0 project page
- SANA-Video 2.0 documentation
- Original 50-step 5B 720p model
- Four-step online preview
Citation
If you use SANA-Video, please cite the SANA-Video work linked from the project page . SANA-Video 2.0-specific citation information will be added when it becomes available.
Runs of Efficient-Large-Model SANA-Video_2.0_5B_720p_4step on huggingface.co
825
Total runs
-2
24-hour runs
-185
3-day runs
-185
7-day runs
-1.7K
30-day runs
More Information About SANA-Video_2.0_5B_720p_4step huggingface.co Model
Url of SANA-Video_2.0_5B_720p_4step
SANA-Video_2.0_5B_720p_4step huggingface.co Url
Provider of SANA-Video_2.0_5B_720p_4step huggingface.co
Other API from Efficient-Large-Model
Updated:December 13 2024
Total runs:
37.1K
Run Growth:
37.1K
Growth Rate:
99.95%
Total runs:
19.1K
Run Growth:
-806
Growth Rate:
-4.23%
Total runs:
5.7K
Run Growth:
5.7K
Growth Rate:
100.00%
Total runs:
3.5K
Run Growth:
2.5K
Growth Rate:
70.97%
Total runs:
3.3K
Run Growth:
-11.2K
Growth Rate:
-338.20%
Total runs:
3.3K
Run Growth:
3.2K
Growth Rate:
97.85%
Total runs:
3.2K
Run Growth:
400
Growth Rate:
12.68%
Total runs:
2.2K
Run Growth:
1.9K
Growth Rate:
85.08%
Total runs:
1.2K
Run Growth:
559
Growth Rate:
46.47%
Total runs:
965
Run Growth:
0
Growth Rate:
0.00%
Total runs:
803
Run Growth:
785
Growth Rate:
97.76%
Total runs:
694
Run Growth:
397
Growth Rate:
57.20%
Total runs:
580
Run Growth:
325
Growth Rate:
56.03%
Total runs:
559
Run Growth:
-2.4K
Growth Rate:
-428.21%
Total runs:
462
Run Growth:
41
Growth Rate:
8.87%
Total runs:
425
Run Growth:
340
Growth Rate:
80.00%
Total runs:
412
Run Growth:
325
Growth Rate:
78.88%
Total runs:
402
Run Growth:
281
Growth Rate:
69.90%
Total runs:
401
Run Growth:
11
Growth Rate:
2.74%
Total runs:
386
Run Growth:
-280
Growth Rate:
-72.54%
Total runs:
376
Run Growth:
362
Growth Rate:
96.28%
Total runs:
357
Run Growth:
313
Growth Rate:
87.68%
Total runs:
338
Run Growth:
338
Growth Rate:
100.00%
Total runs:
295
Run Growth:
-297
Growth Rate:
-100.68%
Total runs:
273
Run Growth:
252
Growth Rate:
93.68%
Total runs:
227
Run Growth:
183
Growth Rate:
80.26%
Total runs:
224
Run Growth:
136
Growth Rate:
60.71%
Total runs:
214
Run Growth:
214
Growth Rate:
100.00%
Total runs:
207
Run Growth:
-350
Growth Rate:
-167.46%
Total runs:
203
Run Growth:
92
Growth Rate:
45.32%
Total runs:
179
Run Growth:
179
Growth Rate:
100.00%
Total runs:
166
Run Growth:
-355
Growth Rate:
-213.86%
Total runs:
152
Run Growth:
141
Growth Rate:
92.76%
Total runs:
145
Run Growth:
24
Growth Rate:
16.55%
Total runs:
118
Run Growth:
0
Growth Rate:
0.00%
Total runs:
117
Run Growth:
91
Growth Rate:
77.78%
Total runs:
117
Run Growth:
-39
Growth Rate:
-33.33%
Total runs:
107
Run Growth:
0
Growth Rate:
0.00%
Total runs:
107
Run Growth:
37
Growth Rate:
34.58%
Total runs:
103
Run Growth:
-15
Growth Rate:
-14.56%
Total runs:
102
Run Growth:
29
Growth Rate:
28.71%
Total runs:
97
Run Growth:
2
Growth Rate:
2.06%
Total runs:
95
Run Growth:
20
Growth Rate:
21.05%
Total runs:
89
Run Growth:
26
Growth Rate:
28.89%
Total runs:
86
Run Growth:
-60
Growth Rate:
-69.77%
Total runs:
86
Run Growth:
21
Growth Rate:
23.86%
Total runs:
76
Run Growth:
13
Growth Rate:
17.11%
Total runs:
76
Run Growth:
76
Growth Rate:
100.00%
Total runs:
73
Run Growth:
-3
Growth Rate:
-4.11%
Total runs:
63
Run Growth:
33
Growth Rate:
52.38%
Total runs:
61
Run Growth:
13
Growth Rate:
21.31%
Total runs:
60
Run Growth:
60
Growth Rate:
100.00%
Total runs:
60
Run Growth:
43
Growth Rate:
100.00%
Total runs:
59
Run Growth:
59
Growth Rate:
100.00%
Total runs:
54
Run Growth:
31
Growth Rate:
100.00%
Total runs:
54
Run Growth:
22
Growth Rate:
40.74%
Total runs:
50
Run Growth:
-24
Growth Rate:
-47.06%
Total runs:
49
Run Growth:
0
Growth Rate:
0.00%
Total runs:
47
Run Growth:
12
Growth Rate:
25.00%
Total runs:
46
Run Growth:
11
Growth Rate:
23.91%
Total runs:
45
Run Growth:
15
Growth Rate:
33.33%
Total runs:
45
Run Growth:
0
Growth Rate:
0.00%
Total runs:
45
Run Growth:
37
Growth Rate:
82.22%
Total runs:
43
Run Growth:
19
Growth Rate:
44.19%
Total runs:
42
Run Growth:
-10
Growth Rate:
-22.73%
Total runs:
40
Run Growth:
-21
Growth Rate:
-52.50%
Total runs:
39
Run Growth:
24
Growth Rate:
61.54%
Total runs:
38
Run Growth:
19
Growth Rate:
50.00%
Total runs:
38
Run Growth:
-8
Growth Rate:
-21.05%
Total runs:
37
Run Growth:
5
Growth Rate:
12.50%
Total runs:
34
Run Growth:
27
Growth Rate:
79.41%
Total runs:
33
Run Growth:
16
Growth Rate:
48.48%
Total runs:
31
Run Growth:
-83
Growth Rate:
-267.74%
Total runs:
27
Run Growth:
0
Growth Rate:
0.00%
Total runs:
27
Run Growth:
15
Growth Rate:
55.56%
Total runs:
27
Run Growth:
20
Growth Rate:
74.07%
Total runs:
25
Run Growth:
19
Growth Rate:
76.00%
Total runs:
25
Run Growth:
14
Growth Rate:
56.00%
Total runs:
24
Run Growth:
24
Growth Rate:
100.00%
Total runs:
24
Run Growth:
9
Growth Rate:
37.50%
Total runs:
23
Run Growth:
9
Growth Rate:
37.50%
Total runs:
23
Run Growth:
14
Growth Rate:
60.87%
Total runs:
21
Run Growth:
-11
Growth Rate:
-52.38%
Total runs:
21
Run Growth:
10
Growth Rate:
47.62%
Total runs:
21
Run Growth:
-1
Growth Rate:
-4.55%
Total runs:
20
Run Growth:
2
Growth Rate:
10.00%
Total runs:
20
Run Growth:
13
Growth Rate:
59.09%
Total runs:
20
Run Growth:
8
Growth Rate:
40.00%
Total runs:
19
Run Growth:
10
Growth Rate:
52.63%
Total runs:
19
Run Growth:
1
Growth Rate:
5.26%
Total runs:
18
Run Growth:
-4
Growth Rate:
-21.05%
Total runs:
18
Run Growth:
-20
Growth Rate:
-111.11%
Total runs:
18
Run Growth:
-5
Growth Rate:
-27.78%
Total runs:
18
Run Growth:
5
Growth Rate:
26.32%
Total runs:
18
Run Growth:
12
Growth Rate:
66.67%
Total runs:
17
Run Growth:
12
Growth Rate:
70.59%
Total runs:
17
Run Growth:
4
Growth Rate:
23.53%
Total runs:
16
Run Growth:
9
Growth Rate:
56.25%
Total runs:
16
Run Growth:
12
Growth Rate:
75.00%