video-to-video
Wan 2.1 V2V 720p LoRA Ultra-Fast converts videos to 720p with custom LoRA support and lets you generate unlimited AI videos. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Idle
Your request will cost $0.225 per run.
For $10 you can run this model approximately 44 times.
One more thing::
ExamplesView all
README
Wan 2.1 V2V 720p LoRA Ultra Fast — wavespeed-ai/wan-2.1/v2v-720p-lora-ultra-fast
Wan 2.1 V2V 720p LoRA Ultra Fast is a speed-optimized video-to-video model that transforms an input video using a text prompt while preserving the original motion and timing. Upload a source video, describe the desired style or changes, and tune strength to balance between “keep the original” and “apply the edit.” It supports up to 3 LoRAs for consistent styling, character look, or branded aesthetics—now with faster turnaround for rapid iteration at 720p.
Key capabilities
- Ultra-fast video-to-video transformation anchored to an input video (720p output)
- Prompt-guided edits while keeping motion continuity and pacing
- Strength control to balance preservation vs. transformation
- LoRA support (up to 3) for stable style/identity steering across clips
- Fine motion behavior tuning via flow_shift
Use cases
- Rapid 720p V2V restyling for social, ads, and creative iteration
- Apply a consistent “house style” across multiple clips using LoRAs
- Upgrade mood and color grade (cinematic, warm window light, neon, noir)
- Brand-safe refresh: keep composition and timing, update textures/colors/details
- Quick A/B testing by changing prompts, LoRAs, or seed
Pricing
| Duration | Price per video |
|---|---|
| 5s | $0.225 |
| 10s | $0.3375 |
Inputs
- video (required): source video to transform
- prompt (required): what to change and how the result should look
- negative_prompt (optional): what to avoid (artifacts, jitter, unwanted elements)
- loras (optional): up to 3 LoRA items for style/identity steering
Parameters
- num_inference_steps: sampling steps
- duration: output duration (seconds)
- strength: how strongly to transform the input video (lower = preserve more; higher = change more)
- guidance_scale: prompt adherence strength
- flow_shift: motion/flow behavior tuning
- seed: random seed (-1 for random; fixed for reproducible results)
LoRA (up to 3 items):
-
loras: list of LoRA entries (max 3)
- path: owner/model-name or a direct .safetensors URL
- scale: LoRA strength
Prompting guide (V2V + LoRA)
A reliable structure is “preserve + edit + style”:
Template: Keep the original motion and timing. Apply [style/look] and adjust [lighting/colors/textures]. Keep faces natural and stable. Avoid flicker, warping, and jitter.
Example prompts
- Keep the original motion and composition. Apply a warm, cozy studio look with soft window light, visible dust particles, gentle film grain, and natural skin tones.
- Preserve camera motion and timing. Restyle the clip into a flat-color illustration look while keeping clean edges and stable shading.
- Keep the scene and movement. Shift the color grade to golden hour, add subtle bloom and soft shadows, maintain realism.