zsxkib/wan-squish – Run with an API on Replicate

zsxkib / wan-squish

Warm

Public
12 runs
H100

Run with an API

Playground API Examples README Versions

Input

prompt

*string

Shift + Return to add a new line

SQUISH-IT Cute golden retriever puppy sitting in grass with flowers. Human hands enter the frame and gently begin to squish and mold the puppy like soft dough. The puppy's fluffy fur and form gradually transform into a malleable clay-like substance as the hands shape it. The final shot shows the reshaped puppy-dough creation sitting on the grass surrounded by flowersSQUISH-IT Cute golden retriever puppy sitting in grass with flowers. Human hands enter the frame and gently begin to squish and mold the puppy like soft dough. The puppy's fluffy fur and form gradually transform into a malleable clay-like substance as the hands shape it. The final shot shows the reshaped puppy-dough creation sitting on the grass surrounded by flowers

Text prompt for video generation

negative_prompt

string

Shift + Return to add a new line

Things you do not want to see in your video

Default: ""

image

file

Image to use as a starting frame for image to video generation.

aspect_ratio

string

The aspect ratio of the video. 16:9, 9:16, 1:1, etc.

Default: "16:9"

frames

integer

The number of frames to generate (1 to 5 seconds)

Default: 81

resolution

string

The resolution of the video. 720p is not supported for 1.3b.

Default: "480p"

lora_strength_model

number

Strength of the LORA applied to the model. 0.0 is no LORA.

Default: 1

lora_strength_clip

number

Strength of the LORA applied to the CLIP model. 0.0 is no LORA.

Default: 1

fast_mode

string

Speed up generation with different levels of acceleration. Faster modes may degrade quality somewhat. The speedup is dependent on the content, so different videos may see different speedups.

Default: "Balanced"

sample_steps

integer

(minimum: 1, maximum: 60)

Number of generation steps. Fewer steps means faster generation, at the expensive of output quality. 30 steps is sufficient for most prompts

Default: 30

sample_guide_scale

number

(minimum: 0, maximum: 10)

Higher guide scale makes prompt adherence better, but can reduce variation

Default: 5

sample_shift

number

(minimum: 0, maximum: 10)

Sample shift factor

Default: 8

seed

integer

Set a seed for reproducibility. Random by default.

Run this model in Node.js with one line of code:

npx create-replicate --model=zsxkib/wan-squish

or set up a project from scratch

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import and set up the client:

import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

Run zsxkib/wan-squish using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "zsxkib/wan-squish:c9eabcc2d6364b071f4a7cc00bc856997d58173c1bd49bae325d44d123f67204",
  {
    input: {
      image: "https://replicate.delivery/pbxt/Mh3wselQD9qkHDtiibSZKUyFBMXNujyCwYMRFFkgY3DcOyx5/Screenshot%202025-03-19%20at%2011.24.14.png",
      frames: 81,
      prompt: "SQUISH-IT Cute golden retriever puppy sitting in grass with flowers. Human hands enter the frame and gently begin to squish and mold the puppy like soft dough. The puppy's fluffy fur and form gradually transform into a malleable clay-like substance as the hands shape it. The final shot shows the reshaped puppy-dough creation sitting on the grass surrounded by flowers",
      fast_mode: "Balanced",
      resolution: "480p",
      aspect_ratio: "16:9",
      sample_shift: 8,
      sample_steps: 30,
      negative_prompt: "",
      lora_strength_clip: 1,
      sample_guide_scale: 5,
      lora_strength_model: 1
    }
  }
);
console.log(output);

To learn more, take a look at the guide on getting started with Node.js.

Install Replicate’s Python client library:

pip install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import the client:

import replicate

Run zsxkib/wan-squish using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "zsxkib/wan-squish:c9eabcc2d6364b071f4a7cc00bc856997d58173c1bd49bae325d44d123f67204",
    input={
        "image": "https://replicate.delivery/pbxt/Mh3wselQD9qkHDtiibSZKUyFBMXNujyCwYMRFFkgY3DcOyx5/Screenshot%202025-03-19%20at%2011.24.14.png",
        "frames": 81,
        "prompt": "SQUISH-IT Cute golden retriever puppy sitting in grass with flowers. Human hands enter the frame and gently begin to squish and mold the puppy like soft dough. The puppy's fluffy fur and form gradually transform into a malleable clay-like substance as the hands shape it. The final shot shows the reshaped puppy-dough creation sitting on the grass surrounded by flowers",
        "fast_mode": "Balanced",
        "resolution": "480p",
        "aspect_ratio": "16:9",
        "sample_shift": 8,
        "sample_steps": 30,
        "negative_prompt": "",
        "lora_strength_clip": 1,
        "sample_guide_scale": 5,
        "lora_strength_model": 1
    }
)
print(output)

To learn more, take a look at the guide on getting started with Python.

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Run zsxkib/wan-squish using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

curl -s -X POST \
  -H "Authorization: Bearer $REPLICATE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Prefer: wait" \
  -d $'{
    "version": "c9eabcc2d6364b071f4a7cc00bc856997d58173c1bd49bae325d44d123f67204",
    "input": {
      "image": "https://replicate.delivery/pbxt/Mh3wselQD9qkHDtiibSZKUyFBMXNujyCwYMRFFkgY3DcOyx5/Screenshot%202025-03-19%20at%2011.24.14.png",
      "frames": 81,
      "prompt": "SQUISH-IT Cute golden retriever puppy sitting in grass with flowers. Human hands enter the frame and gently begin to squish and mold the puppy like soft dough. The puppy\'s fluffy fur and form gradually transform into a malleable clay-like substance as the hands shape it. The final shot shows the reshaped puppy-dough creation sitting on the grass surrounded by flowers",
      "fast_mode": "Balanced",
      "resolution": "480p",
      "aspect_ratio": "16:9",
      "sample_shift": 8,
      "sample_steps": 30,
      "negative_prompt": "",
      "lora_strength_clip": 1,
      "sample_guide_scale": 5,
      "lora_strength_model": 1
    }
  }' \
  https://api.replicate.com/v1/predictions

To learn more, take a look at Replicate’s HTTP API reference docs.

Output

{
  "completed_at": "2025-03-20T18:05:27.060931Z",
  "created_at": "2025-03-20T18:01:42.373000Z",
  "data_removed": false,
  "error": null,
  "id": "5dx795prwnrma0cnpjvrwxzyrr",
  "input": {
    "image": "https://replicate.delivery/pbxt/Mh3wselQD9qkHDtiibSZKUyFBMXNujyCwYMRFFkgY3DcOyx5/Screenshot%202025-03-19%20at%2011.24.14.png",
    "frames": 81,
    "prompt": "SQUISH-IT Cute golden retriever puppy sitting in grass with flowers. Human hands enter the frame and gently begin to squish and mold the puppy like soft dough. The puppy's fluffy fur and form gradually transform into a malleable clay-like substance as the hands shape it. The final shot shows the reshaped puppy-dough creation sitting on the grass surrounded by flowers",
    "fast_mode": "Balanced",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "sample_shift": 8,
    "sample_steps": 30,
    "negative_prompt": "",
    "lora_strength_clip": 1,
    "sample_guide_scale": 5,
    "lora_strength_model": 1
  },
  "logs": "Random seed set to: 4013781049\n2025-03-20T18:01:42Z | INFO  | [ Initiating ] chunk_size=150M dest=/tmp/tmpxsxow2x3/weights url=https://replicate.delivery/xezq/phYRdhAKpYr6B5z5adpBg0gbJRSbInYlmhI4fLdyWlF9UPNKA/trained_model.tar\n2025-03-20T18:01:46Z | INFO  | [ Complete ] dest=/tmp/tmpxsxow2x3/weights size=\"359 MB\" total_elapsed=3.528s url=https://replicate.delivery/xezq/phYRdhAKpYr6B5z5adpBg0gbJRSbInYlmhI4fLdyWlF9UPNKA/trained_model.tar\nChecking inputs\n✅ /tmp/inputs/image.png\n====================================\nChecking weights\n⏳ Downloading wan2.1_i2v_480p_14B_bf16.safetensors to ComfyUI/models/diffusion_models\n✅ wan2.1_i2v_480p_14B_bf16.safetensors downloaded to ComfyUI/models/diffusion_models in 20.14s, size: 31270.88MB\n✅ 14b_0d581762fafe89bfad5f07ae2276cf4e.safetensors exists in loras directory\n✅ umt5_xxl_fp16.safetensors exists in ComfyUI/models/text_encoders\n✅ clip_vision_h.safetensors exists in ComfyUI/models/clip_vision\n✅ wan_2.1_vae.safetensors exists in ComfyUI/models/vae\n====================================\nRunning workflow\n[ComfyUI] got prompt\nExecuting node 55, title: Load Image, class type: LoadImage\nExecuting node 56, title: Width and height for scaling image to ideal resolution 🪴, class type: Width and height for scaling image to ideal resolution 🪴\nExecuting node 57, title: 🔧 Image Resize, class type: ImageResize+\nExecuting node 60, title: Load CLIP Vision, class type: CLIPVisionLoader\n[ComfyUI] Requested to load CLIPVisionModelProjection\nExecuting node 59, title: CLIP Vision Encode, class type: CLIPVisionEncode\n[ComfyUI] loaded completely 129691.98263816834 1208.09814453125 True\nExecuting node 37, title: Load Diffusion Model, class type: UNETLoader\n[ComfyUI] model weight dtype torch.float16, manual cast: None\n[ComfyUI] model_type FLOW\nExecuting node 54, title: WanVideo Tea Cache (native), class type: WanVideoTeaCacheKJ\nExecuting node 49, title: Load LoRA, class type: LoraLoader\n[ComfyUI] Requested to load WanTEModel\nExecuting node 6, title: CLIP Text Encode (Positive Prompt), class type: CLIPTextEncode\n[ComfyUI] loaded completely 139319.35861854552 10835.4765625 True\nExecuting node 58, title: WanImageToVideo, class type: WanImageToVideo\nExecuting node 48, title: ModelSamplingSD3, class type: ModelSamplingSD3\nExecuting node 53, title: WanVideo Enhance A Video (native), class type: WanVideoEnhanceAVideoKJ\nExecuting node 3, title: KSampler, class type: KSampler\n[ComfyUI] Requested to load WAN21\n[ComfyUI] loaded completely 122348.68833397522 31269.802368164062 True\n[ComfyUI]\n[ComfyUI] 0%|          | 0/30 [00:00<?, ?it/s]\n[ComfyUI] 3%|▎         | 1/30 [00:08<03:56,  8.14s/it]\n[ComfyUI] 7%|▋         | 2/30 [00:18<04:23,  9.40s/it]\n[ComfyUI] 10%|█         | 3/30 [00:28<04:24,  9.81s/it]\n[ComfyUI] TeaCache: Initialized\n[ComfyUI]\n[ComfyUI] 13%|█▎        | 4/30 [00:41<04:44, 10.94s/it]\n[ComfyUI] 20%|██        | 6/30 [00:52<03:11,  7.99s/it]\n[ComfyUI] 23%|██▎       | 7/30 [01:02<03:20,  8.72s/it]\n[ComfyUI] 30%|███       | 9/30 [01:13<02:32,  7.25s/it]\n[ComfyUI] 37%|███▋      | 11/30 [01:24<02:03,  6.52s/it]\n[ComfyUI] 43%|████▎     | 13/30 [01:34<01:43,  6.11s/it]\n[ComfyUI] 50%|█████     | 15/30 [01:45<01:27,  5.85s/it]\n[ComfyUI] 57%|█████▋    | 17/30 [01:56<01:14,  5.69s/it]\n[ComfyUI] 63%|██████▎   | 19/30 [02:07<01:01,  5.59s/it]\n[ComfyUI] 70%|███████   | 21/30 [02:17<00:49,  5.52s/it]\n[ComfyUI] 77%|███████▋  | 23/30 [02:17<00:26,  3.84s/it]\n[ComfyUI] 80%|████████  | 24/30 [02:28<00:30,  5.07s/it]\n[ComfyUI] 87%|████████▋ | 26/30 [02:28<00:13,  3.37s/it]\n[ComfyUI] 93%|█████████▎| 28/30 [02:39<00:08,  4.02s/it]\n[ComfyUI] 97%|█████████▋| 29/30 [02:50<00:05,  5.28s/it]\n[ComfyUI] 100%|██████████| 30/30 [03:00<00:00,  6.42s/it]\nExecuting node 8, title: VAE Decode, class type: VAEDecode\nExecuting node 50, title: Video Combine 🎥🅥🅗🅢, class type: VHS_VideoCombine\n[ComfyUI] 100%|██████████| 30/30 [03:00<00:00,  6.03s/it]\n[ComfyUI] Prompt executed in 200.68 seconds\noutputs:  {'50': {'gifs': [{'filename': 'R8_Wan_00001.mp4', 'subfolder': '', 'type': 'output', 'format': 'video/h264-mp4', 'frame_rate': 16.0, 'workflow': 'R8_Wan_00001.png', 'fullpath': '/tmp/outputs/R8_Wan_00001.mp4'}]}}\n====================================\nR8_Wan_00001.png\nR8_Wan_00001.mp4",
  "metrics": {
    "predict_time": 224.680781387,
    "total_time": 224.687931
  },
  "output": [
    "https://replicate.delivery/xezq/A4xWBgbXZbqMJ5f2UsScz4teVKKNTtfA4R4ezIalJcAfcFUjC/R8_Wan_00001.mp4"
  ],
  "started_at": "2025-03-20T18:01:42.380150Z",
  "status": "succeeded",
  "urls": {
    "stream": "https://stream.replicate.com/v1/files/bcwr-my6633lu2o72x6pid6c4l2hfvlj6yrtuuwnep2zuugeeapc6resq",
    "get": "https://api.replicate.com/v1/predictions/5dx795prwnrma0cnpjvrwxzyrr",
    "cancel": "https://api.replicate.com/v1/predictions/5dx795prwnrma0cnpjvrwxzyrr/cancel"
  },
  "version": "c9eabcc2d6364b071f4a7cc00bc856997d58173c1bd49bae325d44d123f67204"
}

Generated in

3 minutes 45 seconds

Tweak it ShareReport

Random seed set to: 4013781049
2025-03-20T18:01:42Z | INFO  | [ Initiating ] chunk_size=150M dest=/tmp/tmpxsxow2x3/weights url=https://replicate.delivery/xezq/phYRdhAKpYr6B5z5adpBg0gbJRSbInYlmhI4fLdyWlF9UPNKA/trained_model.tar
2025-03-20T18:01:46Z | INFO  | [ Complete ] dest=/tmp/tmpxsxow2x3/weights size="359 MB" total_elapsed=3.528s url=https://replicate.delivery/xezq/phYRdhAKpYr6B5z5adpBg0gbJRSbInYlmhI4fLdyWlF9UPNKA/trained_model.tar
Checking inputs
✅ /tmp/inputs/image.png
====================================
Checking weights
⏳ Downloading wan2.1_i2v_480p_14B_bf16.safetensors to ComfyUI/models/diffusion_models
✅ wan2.1_i2v_480p_14B_bf16.safetensors downloaded to ComfyUI/models/diffusion_models in 20.14s, size: 31270.88MB
✅ 14b_0d581762fafe89bfad5f07ae2276cf4e.safetensors exists in loras directory
✅ umt5_xxl_fp16.safetensors exists in ComfyUI/models/text_encoders
✅ clip_vision_h.safetensors exists in ComfyUI/models/clip_vision
✅ wan_2.1_vae.safetensors exists in ComfyUI/models/vae
====================================
Running workflow
[ComfyUI] got prompt
Executing node 55, title: Load Image, class type: LoadImage
Executing node 56, title: Width and height for scaling image to ideal resolution 🪴, class type: Width and height for scaling image to ideal resolution 🪴
Executing node 57, title: 🔧 Image Resize, class type: ImageResize+
Executing node 60, title: Load CLIP Vision, class type: CLIPVisionLoader
[ComfyUI] Requested to load CLIPVisionModelProjection
Executing node 59, title: CLIP Vision Encode, class type: CLIPVisionEncode
[ComfyUI] loaded completely 129691.98263816834 1208.09814453125 True
Executing node 37, title: Load Diffusion Model, class type: UNETLoader
[ComfyUI] model weight dtype torch.float16, manual cast: None
[ComfyUI] model_type FLOW
Executing node 54, title: WanVideo Tea Cache (native), class type: WanVideoTeaCacheKJ
Executing node 49, title: Load LoRA, class type: LoraLoader
[ComfyUI] Requested to load WanTEModel
Executing node 6, title: CLIP Text Encode (Positive Prompt), class type: CLIPTextEncode
[ComfyUI] loaded completely 139319.35861854552 10835.4765625 True
Executing node 58, title: WanImageToVideo, class type: WanImageToVideo
Executing node 48, title: ModelSamplingSD3, class type: ModelSamplingSD3
Executing node 53, title: WanVideo Enhance A Video (native), class type: WanVideoEnhanceAVideoKJ
Executing node 3, title: KSampler, class type: KSampler
[ComfyUI] Requested to load WAN21
[ComfyUI] loaded completely 122348.68833397522 31269.802368164062 True
[ComfyUI]
[ComfyUI] 0%|          | 0/30 [00:00<?, ?it/s]
[ComfyUI] 3%|▎         | 1/30 [00:08<03:56,  8.14s/it]
[ComfyUI] 7%|▋         | 2/30 [00:18<04:23,  9.40s/it]
[ComfyUI] 10%|█         | 3/30 [00:28<04:24,  9.81s/it]
[ComfyUI] TeaCache: Initialized
[ComfyUI]
[ComfyUI] 13%|█▎        | 4/30 [00:41<04:44, 10.94s/it]
[ComfyUI] 20%|██        | 6/30 [00:52<03:11,  7.99s/it]
[ComfyUI] 23%|██▎       | 7/30 [01:02<03:20,  8.72s/it]
[ComfyUI] 30%|███       | 9/30 [01:13<02:32,  7.25s/it]
[ComfyUI] 37%|███▋      | 11/30 [01:24<02:03,  6.52s/it]
[ComfyUI] 43%|████▎     | 13/30 [01:34<01:43,  6.11s/it]
[ComfyUI] 50%|█████     | 15/30 [01:45<01:27,  5.85s/it]
[ComfyUI] 57%|█████▋    | 17/30 [01:56<01:14,  5.69s/it]
[ComfyUI] 63%|██████▎   | 19/30 [02:07<01:01,  5.59s/it]
[ComfyUI] 70%|███████   | 21/30 [02:17<00:49,  5.52s/it]
[ComfyUI] 77%|███████▋  | 23/30 [02:17<00:26,  3.84s/it]
[ComfyUI] 80%|████████  | 24/30 [02:28<00:30,  5.07s/it]
[ComfyUI] 87%|████████▋ | 26/30 [02:28<00:13,  3.37s/it]
[ComfyUI] 93%|█████████▎| 28/30 [02:39<00:08,  4.02s/it]
[ComfyUI] 97%|█████████▋| 29/30 [02:50<00:05,  5.28s/it]
[ComfyUI] 100%|██████████| 30/30 [03:00<00:00,  6.42s/it]
Executing node 8, title: VAE Decode, class type: VAEDecode
Executing node 50, title: Video Combine 🎥🅥🅗🅢, class type: VHS_VideoCombine
[ComfyUI] 100%|██████████| 30/30 [03:00<00:00,  6.03s/it]
[ComfyUI] Prompt executed in 200.68 seconds
outputs:  {'50': {'gifs': [{'filename': 'R8_Wan_00001.mp4', 'subfolder': '', 'type': 'output', 'format': 'video/h264-mp4', 'frame_rate': 16.0, 'workflow': 'R8_Wan_00001.png', 'fullpath': '/tmp/outputs/R8_Wan_00001.mp4'}]}}
====================================
R8_Wan_00001.png
R8_Wan_00001.mp4

Examples

View more examples

Run time and cost

This model runs on Nvidia H100 GPU hardware. We don't yet have enough runs of this model to provide performance information.

Readme

This model doesn't have a readme.