by Alibaba
Lightweight video generation model from Alibaba. Only 1.3B params — runs on consumer GPUs with 8GB+ VRAM. Good quality for its size, excellent for rapid iteration.
Your hardware
Detecting...
Measured quality metrics for Wan Video 2.1 1.3B outputs.
How often humans prefer this model's output (0-100%)
Visual quality and composition rating (5-9 scale)
VRAM estimates at FP16 and FP8 precision. FP8 uses ~40% less memory with minimal quality loss. Grade shows how well each GPU handles the generation workload.
| Scenario | VRAM | RTX 4090 24GB | RTX 3060 12GB | RTX 4060 8GB | MacBook Pro M4 Pro 24GB |
|---|---|---|---|---|---|
| 512×512 · 25 frames | 9.1 GB | S● | S● | B● | S● |
| 768×512 · 25 frames | 9.8 GB | S● | A● | D● | S● |
| 768×512 · 100 frames | 11.8 GB | S● | B● | F● | F● |
| 1280×720 · 25 frames | 12.6 GB | S● | B● | F● | F● |
| Scenario | VRAM | RTX 4090 24GB | RTX 3060 12GB | RTX 4060 8GB | MacBook Pro M4 Pro 24GB |
|---|---|---|---|---|---|
| 512×512 · 25 frames | 16.2 GB | S | D | F | B |
| 768×512 · 25 frames | 18.3 GB | S | F | F | B |
| 768×512 · 100 frames | 24.6 GB | B | F | F | F |
| 1280×720 · 25 frames | 26.7 GB | B | F | F | F |
Turbo / LCM distillation
Use distilled scheduler at 4-8 steps for faster iteration
from diffusers import WanPipeline
import torch
pipe = WanPipeline.from_pretrained(
"Wan-AI/Wan2.1-T2V-1.3B",
torch_dtype=torch.float16
)
pipe.to("cuda")
frames = pipe(
prompt="your prompt here",
num_inference_steps=50,
guidance_scale=5.0,
num_frames=81,
).frames[0]
# Save frames or export as videoGet started
Setup instructions for running Wan Video 2.1 1.3B locally
1. Download the model
Get the checkpoint from HuggingFace
2. Place in:
ComfyUI/models/checkpoints/3. Launch ComfyUI
python main.pyVRAM allocation for 25 frames at 768×512 on RTX 4090 24GB
25 frames at 768×512, 30 steps, FP16.
Download Wan Video 2.1 1.3B in different precisions. Lower precision = less VRAM but slight quality loss.
| Format | Precision | Size | Provider | |
|---|---|---|---|---|
| safetensorsRecommended | FP16 | 2.6 GB | official | Download |
Few LoRAs available for the 1.3B variant.
Frequently asked questions
Wan Video 2.1 1.3B (1.3B parameters) requires approximately 9.8 GB of VRAM at FP16 precision for generating 25 frames at 768×512. Video generation typically requires more VRAM than image generation due to temporal attention layers.
Yes, the RTX 4090 (24 GB VRAM) can run Wan Video 2.1 1.3B at FP16. Expected generation time is around ~1m 13s for a 25-frame clip.
On a reference GPU (RTX 4090 24GB), Wan Video 2.1 1.3B generates a 25-frame video at 768×512 in approximately ~1m 13s at FP16 with 30 inference steps. Faster GPUs with higher memory bandwidth will reduce generation time.
Wan Video 2.1 1.3B supports up to 832×480 resolution and 81 frames per generation at 16 FPS. Higher resolutions and frame counts require proportionally more VRAM.
See also