MotionForge WAN2.2 Fun A14B I2V + LightX2V 4‑Step + Reward LoRAs + 5B Refiner, 32fps

Unlock the full potential of Wan2.2 for creating stunning, high-motion videos from a single image! This efficient and powerful ComfyUI workflow is designed for simplicity and exceptional results. It combines the best available LoRAs and models into a seamless, multi-stage process:

1. Base Generation: Uses the massive Wan2.2-Fun-A14B-InP model with the fantastic Lightx2v 4-step LoRAs for fast, high-quality motion generation.

2. Quality & Style Enhancement: Applies Wan2.2 Reward LoRAs to refine details and artistic style.

3. Latent Upscaling & Refinement: Boosts resolution and cleans up the video using the efficient Wan2.2-Fun-5B-InP model, supercharged by the FastWan LoRA to require only 8 steps.

4. Final Enhancement: Upscales the final video with RealESRGAN_x2plus and interpolates frames with RIFE for a buttery-smooth 32FPS output.

This is a true "set it and forget it" pipeline for achieving professional-looking animated results.

✨ Features & Highlights

• Enhanced Motion: Leverages specialized LoRAs (Lightx2v) designed explicitly for strong, coherent image-to-video motion.

• Two-Pass Quality: Uses both High-Noise and Low-Noise Reward LoRAs for maximum detail and aesthetic quality.

• Efficient HD Upscaling: The 5B upscaler model with the FastWan LoRA delivers high-quality results in just 8 steps, saving you time and VRAM.

• Cinematic Output: Final output is smoothly interpolated to 32FPS and upscaled for a professional finish.

• Organized Workflow: Cleanly grouped nodes make the workflow easy to navigate, understand, and modify.

• GPU Memory Management: Includes cleanGpuUsed nodes to help manage memory during the complex generation process.

📦 Required Models (Please Download First!)

For this workflow to function, you must download and place the following models in your respective ComfyUI models folders.

1. Core Wan2.2 GGUF Models:

• Wan2.2-Fun-A14B-InP_HighNoise-Q8_0.gguf

• Wan2.2-Fun-A14B-InP_LowNoise-Q8_0.gguf

• Wan2.2-Fun-5B-InP-Q8_0.gguf

  • Source: https://huggingface.co/QuantStack/Wan2.2-Fun-A14B-InP-GGUF & https://huggingface.co/QuantStack/Wan2.2-Fun-5B-InP-GGUF

2. Motion & Quality LoRAs (for A14B):

• lightx2v_I2V_14B_480p_cfg_step_distill_rank128_bf16.safetensors (The key to great motion!)

  • Source: https://huggingface.co/Kijai/WanVideo_comfy/tree/main/Lightx2v

• Wan2.2-Fun-A14B-InP-high-noise-HPS2.1.safetensors

• Wan2.2-Fun-A14B-InP-low-noise-HPS2.1.safetensors

  • Source: https://huggingface.co/alibaba-pai/Wan2.2-Fun-Reward-LoRAs/tree/main

3. Upscaler LoRA (for 5B):

• Wan2_2_5B_FastWanFullAttn_lora_rank_128_bf16.safetensors (Drastically reduces required steps!)

  • Source: https://huggingface.co/Kijai/WanVideo_comfy/tree/main/FastWan

4. VAE & Upscaler:

• Wan2_1_VAE_fp32.safetensors (or Wan2.2_VAE.safetensors for the upscaler sub-graph)

• RealESRGAN_x2plus.pth (Standard upscaling model)

  • Source: https://huggingface.co/dtarnow/UPscaler/tree/main (or any standard model repository)

• clip_vision_h.safetensors (For the IP-Adapter-like functionality in the inpaint node)

5. CLIP Encoder:

• umt5-xxl-encoder-Q8_0.gguf (Typically bundled with the Wan GGUF downloads)

⚙️ Installation & Usage

1. Download the Workflow: Download the provided .json file from this Civitai page.

2. Download All Models: Ensure you have all the models listed above downloaded to the correct folders.

3. Load in ComfyUI: Open ComfyUI, drag the .json file into the window, and the workflow will load.

4. Check Loaders: The workflow uses ComfyUI-GGUF and ComfyUI-VideoHelperSuite (VHS). Please ensure you have these custom nodes installed.

5. Input Your Image:

   • Drag your starting image to the "Load START Image" node.

   • (Optional) You can add an ending image to the "Load END Image" node to guide the animation.

6. Modify Prompts: Change the positive and negative prompts in the CLIP Text Encode nodes to suit your image.

7. Queue Prompt! You're ready to go. The workflow will handle the rest, from generation to upscaling and interpolation.

Pro Tip: The default megapixel setting for the upscaler is conservative (0.9). You can increase this value in the "Upscaler" group for higher resolutions, but be mindful of your VRAM!

Conclusion

This workflow represents a curated "best-of" approach to using the Wan2.2 model family. It simplifies a complex multi-step process into a single, powerful pipeline that delivers consistently impressive motion and quality.

Enjoy creating! If you have any questions, feel free to ask in the comments. Don't forget to like and share your amazing results!