Wan 2.1 Synthetic To Real Ditto

Wan 2.1 Synthetic To Real Ditto

Playground

WAN 2.1 Synthetic To Real Ditto mirrors motion and facial expressions in video-to-video synthetic-to-real conversion. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

WAN 2.1 Synthetic-To-Real Ditto — Video-to-Video Model

WAN 2.1 Synthetic-To-Real Ditto converts stylized or synthetic videos (3D, anime, game footage, VTuber-style avatars, etc.) into realistic, live-action–like footage. It mirrors the original motion and facial expressions while replacing the look with a more natural, photographic style. The model is exposed through a ready-to-use REST inference API with fast warm-starts and affordable pricing.

What it does

• Takes a source video with a synthetic or stylized character.
• Tracks body motion, facial expressions, and timing.
• Generates a new realistic human version of the same performance.
• Preserves framing and pacing so you can swap footage without re-editing your timeline.

This makes it ideal for:

• Upgrading animated storyboards into realistic previews
• Turning VTuber or game-style performances into semi-realistic actors
• Rapid prototyping of live-action shots from synthetic previz

Key Capabilities

• High-fidelity motion mirroring Copies head turns, eye blinks, lip movements, and body motion from the input clip with tight temporal alignment.
• Synthetic-to-real translation Transforms toon, 3D, or heavily stylized characters into natural-looking humans while keeping their core identity and staging.
• Consistent lighting and shading Adapts the original scene’s lighting so the new actor feels anchored in the same environment.
• Resolution flexibility Supports both 480p and 720p output for different production needs.

Inputs and Controls

• video (required) Upload or paste the URL of the source video.

• resolution

  • 480p
  • 720p

Pricing

How to Use

1. Upload your synthetic or stylized video in the video field. (up to 120s)
2. Select the desired resolution (480p or 720p).
3. Make sure Enable Safety Checker is checked.
4. Click Run.
5. After processing, preview the real-style output in the right panel and download it for editing or further post-production.

Tips for Best Results

• Use clips with clear, front-facing characters and stable framing to get the best facial detail.
• Avoid heavy motion blur or rapid strobing; clean animation yields more faithful translations.
• Keep clips short when iterating (around 3–5 seconds) to explore different looks quickly and control costs.
• Once you find a look you like, batch-convert key shots from your project to build a consistent, realistic version of your synthetic footage.

# Submit the task
curl --location --request POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/wan-2.1/synthetic-to-real-ditto" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}" \
--data-raw '{
    "resolution": "480p",
    "seed": -1
}'

# Get the result
curl --location --request GET "https://api.wavespeed.ai/api/v3/predictions/${requestId}/result" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}"