After many tests, this is the first and WIP version of what I currently use for Ref2VA generations.Generate LOW RESOLUTION in a first Pass, then upscale and refine when happy with the low quality resolution results.You can generate and see if audio etc works in low Quality (even without upscaling), and still enhance the same seed and video with the 2nd Pass of the workflow.My workflow is optimized (WIP) for 5070ti with 16GB VRAM and 32GB SysRAM.This is WIP so please give me some feedback how it works, what you struggle with and what you think would help others. It's very basic atm, but resulted the best speed without additional nodes etc so far, especially the First Block Cache node is helping a lot with non-Turbo 20 Step generations.Features:<ul><li>Add (currently) up to 3 image references.</li><li>Use 20 base model or 8 step Turbo Lora generation via switch</li><li>Enable or disable the high-resolution upscaling and refinement in the 2nd Pass.</li><li>Keep the same seed to produce small videos first to check for audio quality and composition before enabling the 2nd pass to upscale to higher resolution</li><li>Add your desired Loras, individually to each pass to ensure the Lora weights are applied accordingly, first pass is Ref2VA, most Loras work best with lower strengths but feel free to play around. The 2nd pass uses a FLF2VA model, use Loras here to add details for the high resolution.</li></ul>MODELS<ul><li>Singularity Ref2VA model: <a target="_blank" rel="ugc" href="https://huggingface.co/WarmBloodAban/Minimax-h3_Singularity/blob/main/Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors">https://huggingface.co/WarmBloodAban/Minimax-h3_Singularity/blob/main/Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors</a></li><li>Base FLF2VA / T2V model: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors">https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors</a></li><li>Text Encoder for 50xx Series Nvidia GPUs: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors">https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors</a></li><li>Text Encoder for older GPUs: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/text_encoders/qwen3vl_32b_minimax_h3_int8_convrot.safetensors">https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/text_encoders/qwen3vl_32b_minimax_h3_int8_convrot.safetensors</a></li><li>Video VAE: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_video_vae_int8_convrot.safetensors">https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_video_vae_int8_convrot.safetensors</a></li><li>Audio VAE: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_audio_vae_fp32.safetensors">https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_audio_vae_fp32.safetensors</a></li><li>Ref2VA Turbo Lora: <a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MiniMax-H3-experimental/blob/main/loras/MiniMax-H3-Ref2VA-Acc-8Step_comfy.safetensors">https://huggingface.co/Kijai/MiniMax-H3-experimental/blob/main/loras/MiniMax-H3-Ref2VA-Acc-8Step_comfy.safetensors</a></li><li>Taomate 3 Step Turbo Lora: <a target="_blank" rel="ugc" href="https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI/blob/main/experimental/minimax_h3_taomate_fl2va_3step_ema_comfyui.safetensors">https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI/blob/main/experimental/minimax_h3_taomate_fl2va_3step_ema_comfyui.safetensors</a></li><li>H3 Latent Upscaler: <a target="_blank" rel="ugc" href="https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler">https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler</a></li></ul>Custom Nodes<ul><li><a target="_blank" rel="ugc" href="https://github.com/siraxe/ComfyUI-SA-Nodes-QQ.git">https://github.com/siraxe/ComfyUI-SA-Nodes-QQ.git</a></li><li><a target="_blank" rel="ugc" href="https://github.com/rgthree/rgthree-comfy">https://github.com/rgthree/rgthree-comfy</a></li><li><a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-KJNodes">https://github.com/kijai/ComfyUI-KJNodes</a></li><li><a target="_blank" rel="ugc" href="https://github.com/xmarre/Comfyui_Minimax_h3_latent_Upscaler-Plus">https://github.com/xmarre/Comfyui_Minimax_h3_latent_Upscaler-Plus</a></li><li><a target="_blank" rel="ugc" href="https://github.com/duckyshell/ComfyUI-MiniMaxH3-FirstBlockCache">https://github.com/duckyshell/ComfyUI-MiniMaxH3-FirstBlockCache</a></li></ul>

MMH3 Ref2VA - FAST 2 Pass Workflow (WIP - Need Feedback)

First official version for testing and feedback from you guys

Files with SHA256: 7487da8872e9223ebf3287c36317056e143aa02ad09acf30c57bae9c60fa636c