Wan I2V 480p LoRA Ultra Fast

Wan-2.1 i2v model with LoRA, generate high-quality videos with superior visual quality and motion diversity

Features

Wan-2.1/i2v-480p-lora-ultra-fast is an open-source AI video generation model developed by Alibaba Cloud, designed for image-to-video tasks.The 14-billion-parameter professional version excels in generating complex motions and simulating physical dynamics, delivering exceptional performance. LoRA stands for Low-Rank Adaptation, a technique for efficiently fine-tuning pre-trained models to generate videos with specified effects from reference images.

Built upon a causal 3D Variational Autoencoder (VAE) and Video Diffusion Transformer architecture, Wan-2.1/i2v-480p-lora-ultra-fast efficiently models spatiotemporal dependencies. In the authoritative VBench evaluation, the 14B version achieved a leading score of 86.22%, surpassing models like Sora, Luma, and Pika, and securing the top position. The model is available on Wavespeed AI, providing convenient access for developers.Leveraging cutting-edge acceleration techniques, Wan-2.1/i2v-480p-lora-ultra-fast pushes the limits of rapid video synthesis for creative and practical applications.

Key Features

• LoRA-Enhanced Stylization: Incorporates LoRA modules to facilitate fine-grained control over stylistic elements, enabling users to generate videos with specific artistic or thematic styles.
• Ultra-Fast Inference: Optimized for speed, this model significantly reduces video generation time compared to standard versions, making it suitable for real-time applications.
• High-Quality 480p Output: Generates sharp and visually appealing 480p videos, maintaining motion diversity and temporal consistency.
• Consumer-Grade GPU Compatibility: Designed to run efficiently on widely available hardware, requiring approximately 8.19 GB of VRAM, thus accessible to a broad range of users.
• Robust Video VAE Integration: Utilizes a powerful variational autoencoder to encode and decode video content effectively, ensuring quality and coherence in the generated outputs.

ComfyUI

Wan-2.1/i2v-480p-lora-ultra-fast is compatible with ComfyUI, offering a node-based workflow for local inference. This integration allows users to customize their video generation processes flexibly and efficiently on their systems.

Limitations

• Creative Focus: The model is intended for creative video synthesis and may not produce factually accurate or reliable content.
• Inherent Biases: As with any data-driven model, outputs may reflect biases present in the training data.
• Input Sensitivity: The quality and consistency of generated videos depend significantly on the quality of the input image; subtle variations may lead to output variability.
• Resolution Limitation: This model is optimized for 480p video generation and does not support higher resolutions like 720p or 1080p.

Out-of-Scope Use

The model and its derivatives may not be used in any way that violates applicable national, federal, state, local, or international law or regulation, including but not limited to:

• Exploiting, harming, or attempting to exploit or harm minors, including solicitation, creation, acquisition, or dissemination of child exploitative content.
• Generating or disseminating verifiably false information with the intent to harm others.
• Creating or distributing personal identifiable information that could be used to harm an individual.
• Harassing, abusing, threatening, stalking, or bullying individuals or groups.
• Producing non-consensual nudity or illegal pornographic content.
• Making fully automated decisions that adversely affect an individual’s legal rights or create binding obligations.
• Facilitating large-scale disinformation campaigns.

Accelerated Inference

Our accelerated inference approach leverages advanced optimization technology from WavespeedAI. This innovative fusion technique significantly reduces computational overhead and latency, enabling rapid image generation without compromising quality. The entire system is designed to efficiently handle large-scale inference tasks while ensuring that real-time applications achieve an optimal balance between speed and accuracy. For further details, please refer to the blog post.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

# Submit the task
curl --location --request POST "https://api.wavespeed.ai/api/v2/wavespeed-ai/wan-2.1/i2v-480p-lora-ultra-fast" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}" \
--data-raw '{
    "image": "https://d2g64w682n9w0w.cloudfront.net/media/images/1743000120316596477_Cea6zwsp.png",
    "prompt": "Felt doll spinning on grassy field, r0t4tion 360 degrees rotation",
    "negative_prompt": "",
    "loras": [
        {
            "path": "Remade-AI/Rotate",
            "scale": 1
        }
    ],
    "size": "832*480",
    "num_inference_steps": 30,
    "duration": 5,
    "guidance_scale": 5,
    "flow_shift": 3,
    "seed": -1,
    "enable_safety_checker": true
}'

# Get the result
curl --location --request GET "https://api.wavespeed.ai/api/v2/predictions/${requestId}/result" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}"

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Query Parameters

Result Request Parameters

Result Response Parameters