lucataco/wan2.1-4step 🔢📝🖼️ → 🖼️

▶️ 4.4K runs 📅 Jul 2025 ⚙️ Cog 0.15.10 🔗 GitHub ⚖️ License

image-to-video

About

Wan 2.1 I2V (14B) 480p with CausVid LoRA

Example Output

Prompt:

"a black horse galloping across a desert"

Output

Performance Metrics

21.28s Prediction Time

260.93s Total Time

All Input Parameters

{
  "seed": -1,
  "steps": 4,
  "prompt": "a black horse galloping across a desert",
  "num_frames": 81,
  "input_image": "https://replicate.delivery/pbxt/NKQD3gyTJ8GgTl6oHVXPaOKR6kr9iiVMs0RMQs86bQYOkaq7/replicate-prediction-phq4vkntgxrme0cqxz0a9amv5w.jpg",
  "guidance_scale": 1,
  "negative_prompt": "",
  "frames_per_second": 16
}

Input Parameters

seed Type: integerDefault: -1: Random seed for reproducible results. Set to -1 for random seed
steps Type: integerDefault: 4Range: 1 - 30: Number of inference steps. More steps = higher quality but slower
width Type: integerRange: 128 - 896: Output width (will be auto-calculated if not provided, must be multiple of 32)
height Type: integerRange: 128 - 896: Output height (will be auto-calculated if not provided, must be multiple of 32)
prompt Type: stringDefault: make this image come alive, cinematic motion, smooth animation: Text prompt describing the desired animation or motion
num_frames Type: integerDefault: 81Range: 2 - 81: Number of frames to generate
input_image (required) Type: string: Input image to animate
guidance_scale Type: numberDefault: 1Range: 0 - 20: Controls adherence to the prompt. Higher values = more adherence
negative_prompt Type: stringDefault:: Negative prompt to avoid unwanted elements
frames_per_second Type: integerDefault: 16Range: 5 - 24: Frames per second for the output video

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Seed: 58013
Generating video: 832x448, 81 frames, 5.06s at 16fps
Using generator device: cuda
  0%|          | 0/4 [00:00<?, ?it/s]
 25%|██▌       | 1/4 [00:04<00:13,  4.34s/it]
 50%|█████     | 2/4 [00:08<00:08,  4.32s/it]
 75%|███████▌  | 3/4 [00:12<00:04,  4.32s/it]
100%|██████████| 4/4 [00:17<00:00,  4.32s/it]
100%|██████████| 4/4 [00:17<00:00,  4.32s/it]
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)

Version Details

Version ID: e3e2b581dffc5a971ab8ef6322f53d93b83c277ec802a5d6bae0f3b62cf592bf
Version Created: July 11, 2025

Run on Replicate →