by Wan-AI
5B text+image-to-video model from the Wan 2.2 family. Runs on consumer GPUs with 8GB+ VRAM. Takes text and reference image as input to generate coherent video clips.
Your hardware
Detecting...
Measured quality metrics for Wan2.2 TI2V 5B 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 | 25.3 GB | B | F | F | F |
| 768×512 · 25 frames | 27.4 GB | B | F | F | F |
| 768×512 · 100 frames | 33.7 GB | F | F | F | F |
| 1280×720 · 25 frames | 35.9 GB | F | F | F | F |
| Scenario | VRAM | RTX 4090 24GB | RTX 3060 12GB | RTX 4060 8GB | MacBook Pro M4 Pro 24GB |
|---|---|---|---|---|---|
| 512×512 · 25 frames | 15.1 GB | S | D | F | A |
| 768×512 · 25 frames | 17.2 GB | S | F | F | B |
| 768×512 · 100 frames | 23.5 GB | B | F | F | D |
| 1280×720 · 25 frames | 25.7 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/Wan2.2-TI2V-5B-Diffusers",
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 Wan2.2 TI2V 5B 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 Wan2.2 TI2V 5B in different precisions. Lower precision = less VRAM but slight quality loss.
| Format | Precision | Size | Provider | |
|---|---|---|---|---|
| safetensorsRecommended | BF16 | 10.5 GB | official | Download |
Early LoRA ecosystem. Compatible with some Wan 2.1 LoRAs.
Frequently asked questions
Wan2.2 TI2V 5B (5B parameters) requires approximately 27.4 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.
Wan2.2 TI2V 5B can run on the RTX 4090 with sequential offloading, though video generation will be significantly slower than native fit.
On a reference GPU (RTX 4090 24GB), Wan2.2 TI2V 5B generates a 25-frame video at 768×512 in approximately ~4m 23s at FP16 with 30 inference steps. Faster GPUs with higher memory bandwidth will reduce generation time.
Wan2.2 TI2V 5B 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