veo-3-1/reference-to-video
Create rapid high-quality video drafts with precise style and speed
| Parameter | Type | Default/Range | Description |
|---|---|---|---|
| prompt | string | "" | Natural-language instruction describing the edit to apply to the input video. |
| video_url | string (video_uri) | "" | URL to the source video to be restyled. Supported formats typically include MP4/MOV/WEBM/GIF. |
| Parameter | Type | Default/Range | Description |
|---|---|---|---|
| resolution | string (enum) | 720p | Output resolution selector. Lucy Restyle currently supports 720p. |
Developers can seamlessly integrate Lucy Restyle using the RunComfy API with standard HTTP requests. The minimal parameter set (prompt, video_url) and predictable 720p output streamline automation in CI/CD pipelines and cloud render farms.
Note: API Endpoint for Lucy Restyle
Create rapid high-quality video drafts with precise style and speed
Create fluid, expressive animations with multi-shot storytelling features.
Lifelike characters, realistic physics, and stunning effects.
Create dynamic, sound-synced motion clips from visuals for rich storytelling.
Turn text prompts into high quality videos with Tencent Hunyuan Video.
Transform visuals into smooth 4K motion clips with sync audio and rapid rendering.
Lucy Restyle video-to-video enables users to transform existing footage using text prompts — changing outfits, backgrounds, lighting, or character styles — while retaining motion consistency and identity integrity. It specializes in restyling, making it ideal for rapid creative iteration or commercial video enhancements.
Lucy Restyle video-to-video focuses explicitly on editing existing videos with high motion stability, while Wan 2.5 or similar models often mix text-to-video generation. Though Lucy Restyle outputs at lower resolution (up to 720p Pro), it excels at preserving realism, facial identity, and coherence during transformations.
Lucy Restyle video-to-video currently outputs up to 720p resolution in the Pro variant and up to 480p in Fast or Dev modes. Supported aspect ratios are generally fixed to match input footage, most commonly 16:9 or 9:16. Custom aspect ratios may yield cropped or letterboxed results.
Lucy Restyle video-to-video typically supports one primary video input plus a text prompt. Advanced modes may allow one additional reference input (e.g., via IP-Adapter or ControlNet). However, multiple layered references can reduce temporal stability, so one-to-one prompt pairing is recommended.
After validating results in RunComfy’s web Playground, developers can use the Lucy Restyle video-to-video endpoint via the RunComfy API by matching model IDs and prompt parameters. You’ll authenticate with API keys, manage USD credit usage, and automate batch editing while keeping outputs consistent with playground behavior.
Lucy Restyle video-to-video Pro provides higher-fidelity edits with excellent identity retention and stability, plus licensing for commercial use. It's more suitable for production pipelines or client projects than Dev or Fast, but users should still review Decart’s original license for specific commercial conditions.
Yes. The Lucy Restyle video-to-video Live configuration supports near-realtime operations, with sub-100ms editing latency in some modes. This makes it ideal for streamers or XR publishers seeking instant restyling effects directly within live broadcasts.
Lucy Restyle video-to-video employs temporal consistency tracking, ensuring visual coherence from frame to frame. It preserves posture, lighting, and facial landmarks, critical for scenes where subjects move dynamically or where continuity between frames must remain intact.
Lucy Restyle video-to-video performs best for localized style modifications—like costume swaps or background restyling—rather than total scene regeneration. Large-scale replacements may cause minor artifacts or drift, particularly in the Dev variant. Use precise prompts for clean transitions.
Lucy Restyle video-to-video supports up to around 512–768 tokens per text prompt depending on the variant. While longer prompts are truncated, concise, directive phrasing (e.g., 'turn actor into astronaut in desert at sunset') tends to deliver more consistent results.
RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.
