Wan3.0: Image to Video

Create up to 30 seconds of high-fidelity video from a single image using Wan 3.0, which supports text, image, audio, and document inputs for versatile scene generation. Produces authentic, realistic footage with unique facial details and believable motion in one pass. Ideal for cinematic storytelling, product visualization, and documentary-style scene recreation.

Download Workflow Try in Cloud

Input

First frame

Prompt

Generate audio

Duration (seconds)

Aspect ratio

Resolution

Advanced settings

Prompt extend

Seed

Run

Output

00:00

Example

Call this model from your code

The request below uses the same prepared graph and settings as the Playground. Each submission can start a paid run; do not automatically retry a submission whose outcome is unknown.

Get API key

What you need

Runs on: Comfy Cloud
Key: COMFY_API_KEY
Your files: Uploaded by the code from their urls

Needs a paid Cloud plan and credits.

Download API JSON

# Python 3.10+: pip install comfy-sdk==0.4.0
import os

from comfy_sdk import Comfy

api_key = os.environ.get("COMFY_API_KEY", "").strip()
if not api_key: raise RuntimeError("Set COMFY_API_KEY")
os.environ.setdefault("COMFY_BASE_URL", "https://cloud.comfy.org")
client = Comfy(api_key=api_key)
workflow = client.workflows.from_json({
    "5": {
        "class_type": "Wan3ImageToVideoApi",
        "inputs": {
            "first_frame": [
                "6",
                0
            ],
            "model": "wan3.0-video",
            "model.audio": True,
            "model.duration": "5",
            "model.prompt": "The midnight-blue rally coupe slides sideways across the old stone viaduct, its rear swinging out as the driver counter-steers hard, then the tyres bite and the car straightens and accelerates out of frame toward the far end of the bridge. A huge curtain of amber dust and shale spray blooms from the rear wheels and drifts over the parapet into the misty canyon. The camera whip-pans to follow the car from a low angle beside the balustrade, then rises in a slow crane move to reveal the river and pine forest far below. Sound of a howling engine, gravel spitting against stone, a turbo whistle and wind in the canyon. No text, no music.",
            "model.prompt_extend": True,
            "model.ratio": "16:9",
            "model.resolution": "720P",
            "seed": 1032549103,
            "watermark": False
        },
        "_meta": {
            "title": "Wan 3.0 Image to Video"
        }
    },
    "6": {
        "class_type": "LoadImage",
        "inputs": {
            "image": "UPLOADED_image_FILENAME"
        },
        "_meta": {
            "title": "Load Image"
        }
    },
    "8": {
        "class_type": "SaveVideo",
        "inputs": {
            "filename_prefix": "video/Wan3.0",
            "format": "auto",
            "format.codec": "auto",
            "video": [
                "5",
                0
            ]
        },
        "_meta": {
            "title": "Save Video"
        }
    }
})
workflow.set_input("6", "image", client.assets.from_url("https://comfy.org/catalog-media/ff3653bd-3dc9-4f80-bb50-b24550fd260f/24f6f711-487b-452c-8db7-11e7eb04af80"))

# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)
for output in job.get_outputs("8"):
    output.to_file(output.name)

API documentation

Try an example

About this workflow

This ComfyUI workflow turns a single still image into a high-fidelity video clip using the Wan 3.0 Image-to-Video API. It’s built around four nodes: LoadImage to bring your source frame into the graph, Wan3ImageToVideoApi to generate the animated sequence on a remote service, SaveVideo to encode the frames to a standard video file, and MarkdownNote for inline instructions. The pipeline is simple and reliable: your input image is loaded, sent to Wan 3.0 for temporal synthesis, and the returned frames are packaged into a playable video.

Technically, Wan 3.0 performs motion synthesis conditioned on the content of the single image, producing realistic camera/lens movement and subject dynamics in one pass. While the Wan 3.0 family supports multimodal inputs, this specific workflow is image-only and focuses on up to 30 seconds of output. You control the key generation settings in the Wan3ImageToVideoApi node (e.g., clip duration and any parameters exposed by your API endpoint), and finalize playback characteristics like FPS and file path in SaveVideo.

This makes it well-suited for cinematic b-roll, product hero shots, and documentary-style scene recreation directly from a still.

Wan3.0: Image to Video

Input

Inside the workflow

Call this model from your code

What you need

Try an example

About this workflow

Questions & answers

Input

Inside the workflow

Call this model from your code

What you need

Try an example

About this workflow

Questions & answers

Do I need an API key for the Wan3ImageToVideoApi node?

What is the maximum video length this workflow can generate?

Which settings control smoothness and playback speed?

Can I add text or audio prompts in this workflow?