<h3 id="wan-2.2-14b-i2v-and-t2v-enhanced-motion-+-5b-latent-upscaler-ultimate-6-steps-hd-pipeline">Wan 2.2 14B i2v &amp; t2v Enhanced Motion + 5B Latent Upscaler - Ultimate 6 Steps HD Pipeline</h3>The ONLY workflow you need. Fixes slow motion, boosts detail with Pusa LoRAs, and features a revolutionary 2-stage upscaler with WanNAG for breathtaking HD videos. Just load your image and go!<hr /><h3 id="the-ultimate-wan2.2-workflow-is-here!-tired-of-these-problems">🚀 The Ultimate Wan2.2 Workflow is HERE! Tired of these problems?</h3><ul><li>Slow, sluggish motion from your Wan2.2 generations?</li><li>Low-quality, blurry results when you try to generate faster?</li><li>VRAM errors when trying to upscale to HD?</li><li>Complex, messy workflows that are hard to manage?</li></ul>This all-in-one solution fixes it ALL. We've cracked the code on high-speed, high-motion, high-detail generation.This isn't just another workflow; it's a complete, optimized production pipeline that takes you from a single image to a stunning, smooth, high-definition video with unparalleled ease and efficiency. Everything is automated and packaged in a clean, intuitive interface using subgraphs for a clutter-free experience.<hr /><h3 id="revolutionary-features-and-&quot;magic-sauce&quot;-ingredients:">✨ Revolutionary Features &amp; "Magic Sauce" Ingredients:</h3>1. 🎯 AUTOMATED &amp; USER-FRIENDLY<ul><li>Fully Automatic Scaling: Just plug in your image! The workflow intelligently analyzes and scales it to the perfect resolution (~0.23 Megapixels) for the Wan 14B model, ensuring optimal stability and quality without any manual input.</li><li>Clean, Subgraph Architecture: The complex tech is hidden away in organized, collapsible groups ("Settings", "Prompts", "Upscaler"). What you see is a simple, linear flow: Image -&gt; Prompts -&gt; SD Output -&gt; HD Output. It’s powerful, but not complicated.</li></ul>2. ⚡ ENHANCED MOTION ENGINE (The 14B Core)This is the heart of the solution. We solve the slow-motion problem with a sophisticated dual-sampler system:<ul><li>Dual Model Power: Uses both the Wan2.2-I2V-A14B-HighNoise and -LowNoise models in tandem.</li><li>Pusa LoRA Quality Anchor: The breakthrough! We inject Pusa V1 LoRAs (<code>HIGH_resized</code> @ 1.5, <code>LOW_resized</code> @ 1.4) into both models. This allows us to run at an incredibly low 6 steps while preserving the sharp details, contrast, and texture of a high-step generation. No more quality loss for speed!</li><li>Lightx2v Motion Catalyst: To supercharge motion at low steps, we apply the powerful lightx2v 14B LoRA at different strengths: a massive 5.6 strength on the High-Noise model to establish strong, coherent motion, and a refined 2.0 strength on the Low-Noise model to clean it up. Result: Dynamic motion without the slowness.</li></ul>3. 🎨 LOW-RAM HD UPsCALING CHAIN (The 5B Power-Up)This is where your video becomes a masterpiece. A genius 2-stage process that is shockingly light on VRAM:<ul><li>Stage 1 - RealESRGAN x2: The initial video is first upscaled 2x for a solid foundation.</li><li>Stage 2 - Latent Detail Injection: This is the secret weapon. The upscaled frames are refined in the latent space by the Wan2.2-TI2V-5B model.<ul><li>FastWan LoRA: We use the FastWanFullAttn LoRA to make the 5B model efficient, requiring only 6 steps at a denoise of 0.2.</li><li>WanVideoNAG Node: Critically, this stage uses the WanVideoNAG (Nested Adaptive Gradient) technique. This allows us to use a very low CFG (1.0) for natural, non-burned images while maintaining the power of your negative prompt to eliminate artifacts and guide the upscale. It’s the best of both worlds.</li></ul></li><li>Result: You get the incredible detail and coherence of a 5B model pass without the typical massive VRAM cost.</li></ul>4. 🍿 CINEMATIC FINISHING TOUCHES<ul><li>RIFE Frame Interpolation: The final step. The upscaled video is interpolated to a silky-smooth 32 FPS, eliminating any minor stutter and delivering a professional, cinematic motion quality.</li></ul><hr /><h3 id="technical-summary-and-requirements:">📊 Technical Summary &amp; Requirements:</h3><ul><li>Core Tech: Advanced dual KSamplerAdvanced setup, Latent Upscaling, WanNAG, RIFE VFI.</li><li>Steps: Only 6 steps for both 14B generation and 5B upscaling.</li><li>Output: Two auto-saved videos: Initial SD (640x352@16fps) and Final HD (1280x704@32fps).</li><li>Optimization: Includes Patch Sage Attention, Torch FP16 patches, and automatic GPU RAM cleanup for maximum stability.</li></ul>Required Models (GGUF Format):<ul><li>14B Models: <code>Wan2.2-I2V-A14B-HighNoise-Q8_0.gguf</code> &amp; <code>LowNoise-Q8_0.gguf</code></li><li>5B Model: <code>Wan2.2-TI2V-5B-Q8_0.gguf</code></li><li>LoRAs: Pusa (High/Low), Lightx2v 14B, FastWan 5B</li><li>Text Encoder: <code>umt5-xxl-encoder-Q8_0.gguf</code></li><li>VAE: <code>Wan2_1_VAE_fp32.safetensors</code> or <code>pig_wan2_vae_fp32-f16.gguf</code></li><li>Upscaler: <code>RealESRGAN_x2plus.pth</code></li></ul>Required Custom Nodes: VideoHelperSuite, Frame-Interpolation, KJNodes, easy-use, gguf, memory_cleanup.<hr /><h3 id="how-to-use-(it's-simple!):">🎬 How to Use (It's Simple!):</h3><ol><li>DOWNLOAD the workflow and all models (links below).</li><li>DRAG &amp; DROP the <code>.json</code> file into ComfyUI.</li><li>CLICK on the "Load Image" node to choose your input picture.</li><li>EDIT the prompts in the "CLIP Text Encode" nodes. The positive prompt includes detailed motion instructions – make it your own!</li><li>QUEUE PROMPT and watch the magic unfold.</li></ol>That's it! The workflow handles everything else automatically.<hr /><h3 id="download-links">🔗 Download Links</h3>Place these in your <code>ComfyUI/models/</code> directory in the corresponding folders.Model / LoRA / ComponentWan2.2 14B - High Noise<a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main">Wan2.2-I2V-A14B-HighNoise-Q8_0.gguf</a>Wan2.2 14B - Low Noise<a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main">Wan2.2-I2V-A14B-LowNoise-Q8_0.gguf</a>Wan2.2 5B - TI2V<a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main">Wan2.2-TI2V-5B-Q8_0.gguf</a>Pusa V1 LoRA (High)<a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/tree/main/Pusa">Wan22_PusaV1_lora_HIGH_resized_dynamic_avg_rank_98_bf16.safetensors</a>Pusa V1 LoRA (Low)<a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/tree/main/Pusa">Wan22_PusaV1_lora_LOW_resized_dynamic_avg_rank_98_bf16.safetensors</a>Lightx2v 14B LoRA<a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/tree/main/Lightx2v">lightx2v_I2V_14B_480p_cfg_step_distill_rank128_bf16.safetensors</a>FastWan 5B LoRA<a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/WanVideo_comfy/tree/main/FastWan">Wan2_2_5B_FastWanFullAttn_lora_rank_128_bf16.safetensors</a>UMT5-XXL Text Encoder<a target="_blank" rel="ugc" href="https://huggingface.co/city96/umt5-xxl-encoder-gguf/tree/main">umt5-xxl-encoder-Q8_0.gguf</a>Wan VAE<a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main/VAE">Wan2_1_VAE_fp32.safetensors</a> / <a target="_blank" rel="ugc" href="https://huggingface.co/QuantStack/Wan2.2-I2V-A14B-GGUF/tree/main/VAE">pig_wan2_vae_fp32-f16.gguf</a>RealESRGAN Upscaler<a target="_blank" rel="ugc" href="https://huggingface.co/ai-forever/Real-ESRGAN/tree/main">RealESRGAN_x2plus.pth</a>RIFE Model<code>rife49.pth</code> (Usually bundled with ComfyUI-Frame-Interpolation)Required Custom Nodes:<ul><li>ComfyUI-VideoHelperSuite</li><li>ComfyUI-Frame-Interpolation</li><li>ComfyUI-KJNodes</li><li>ComfyUI-easy-use</li><li>ComfyUI-gguf (or ComfyUI-GGUF-Integrated)</li><li>ComfyUI_memory_cleanup</li></ul> Transform your ideas into fluid, high-definition reality. Download now and experience the future of Wan2.2 video generation!

Files with SHA256: 151155e354152605064a330c2ddf156a269f21a839e8373a10f1c33e9279c265