<h2 id="what-this-is">What this is?</h2><p>My H3 workflow, designed to have quick generations on my relatively low VRAM (12gb 3060).</p><p>On my setup, this workflow has the following generation times (compared to a basic 4 step workflow), rounded to the nearest 30 seconds:</p><pre><code>5 seconds, I2V
| Megapixels: | Basic 4 Step | V1.0   |
---------------------------------------
| 0.5         | 3:30         | 2:30   |
| 1.0         | 9:30         | 5:30   |
| 2.0         | 30:00        | 13:30  |

10 seconds, I2V
| Megapixels: | Basic 4 Step | V1.0   |
---------------------------------------
| 0.5         | 9:00         | 4:30   |
| 1.0         | 27:30        | 10:00  |
| 2.0         | OOM, DNF     | 20:30  |</code></pre><h2 id="features">Features</h2><h3 id="supports-both-ref2va-and-fl2va">Supports both ref2va and fl2va</h3><p>Select a fl2va model and toggle the 'use fl2va?' option to run in fl2va mode, leave it off and load a ref2va model for ref2va mode.</p><h3 id="automatically-constructs-prompts">Automatically constructs prompts</h3><p>Each section of a H3 prompt (subject_descriptions, retention_analysis, etc.) has it's own input, all of which are combined to construct the prompt, making it easier to keep track of everything.</p><h3 id="dynamic-prompts">Dynamic prompts</h3><p>Uses the <a target="_blank" rel="ugc" href="https://github.com/exectails/comfyui-et_dynamicprompts">dynamic prompts extension</a> to do some pretty neat things, the most useful in this case is using variables in your prompt.</p><h3 id="lora-support">LoRA Support</h3><p>Supports any LoRAs you want. Note: Adding LoRAs will increase generation time.</p><h3 id="easy-references">Easy references</h3><p>Supports the maximum amount of references ref2va uses (9 image, 3 video and 3 audio), and will automatically disable image references 9, 8 and 7 for each reference over the maximum of 12.</p><h3 id="resize-references">Resize references</h3><p>Allows the resizing of image and video references to match the selected generation size, useful for fl2va and keyframes.</p><h2 id="required-node-packs">Required node packs</h2><p>Uses the following node packs:</p><pre><code>🟢 Will be auto-detected by comfyui-manager
🟡 Can be installed using comfyui-manager, but is not auto-detected (you will have to search for it). Manual installation may be easier to avoid getting the wrong pack.
🔴 Requires manual installation</code></pre><p>🟢 <a target="_blank" rel="ugc" href="https://github.com/ltdrdata/ComfyUI-Impact-Pack">Impact Pack</a></p><p>🟢 <a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-KJNodes">KJ Nodes</a></p><p>🟢 <a target="_blank" rel="ugc" href="https://github.com/rgthree/rgthree-comfy">RGThree Comfy</a></p><p>🟡 <a target="_blank" rel="ugc" href="https://github.com/exectails/comfyui-et_dynamicprompts">Dynamic Prompts</a></p><p>🔴 <a target="_blank" rel="ugc" href="https://github.com/T8mars/comfyui-minimax-h3-audio-T8">MiniMax H3 Audio T8</a> Note: This one can be installed using comfyui-manager (it is not auto-detected), but reports it's compatibility requirements incorrectly so it cannot easily be enabled. Manual installation bypasses that.</p><p>🔴 <a target="_blank" rel="ugc" href="https://github.com/yichengup/ComfyUI-YCNodes-MiniMax-H3">YC Nodes MiniMax H3</a></p><p>🔴 <a target="_blank" rel="ugc" href="https://github.com/slmonker/selflift-Avatar">Selflift Avatar</a></p><h2 id="required-models">Required models</h2><h3 id="diffusion-model">Diffusion model</h3><p>Any int8-convrot model, I use <a target="_blank" rel="ugc" href="https://huggingface.co/smhfacct/Minimax-H3-fl2va-ref2va-hybrid-models">these hybrid models</a> for ref2va, and the <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors">basic fl2va one</a>. These go in models/diffusion_models</p><h3 id="text-encoder">Text encoder</h3><p><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors">This one</a>. Goes in models/text_encoders</p><h3 id="vae">Vae</h3><p><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_video_vae_int8_convrot.safetensors">Video</a>, <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_audio_vae_fp32.safetensors">audio</a>. Goes in models/vae</p><h3 id="latent-upscaler">Latent upscaler</h3><p><a target="_blank" rel="ugc" href="https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler/blob/main/minimax_h3_latent_upscaler_3d_conv_v1/minimax_h3_latent_upscaler_3d_conv_v1_bf16.safetensors">This one</a>. Goes in models/latent_upscale_models</p><h3 id="4-step-lora">4 Step LoRA</h3><p>Any pruned LoRA model (must be pruned, will crash otherwise). I use <a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MiniMax-H3-experimental/blob/main/loras/minimax_h3_4step_lora_flashgen_v1.0_768p_fl2va_pruned_avg_rank_13_bf16.safetensors">this one</a>, <a target="_blank" rel="ugc" href="https://civarchive.com/models/2837571/minimax-h3-turbo-loras?modelVersionId=3202732">this one</a> is also pretty good. Goes in models/loras</p><h3 id="full-model-paths">Full model paths</h3><pre><code>📂 ComfyUI
├── 📂 models
|   ├── 📂 diffusion_models
|   |    ├── DIFFUSION MODELS GO HERE
|   |
|   ├── 📂 latent_upscale_models
|   |    ├── LATENT UPSCALER GOES HERE
|   |
|   ├── 📂 loras
|   |    ├── 4 STEP LORA GOES HERE
|   |
|   ├── 📂 text_encoders
|   |    ├── TEXT ENCODER GOES HERE
|   |
|   ├── 📂 vae
|   |    ├── VAE MODELS GO HERE</code></pre><p></p>

MiniMax H3 ref2va/fl2va fast generation workflow

<p>First version I'm happy with. I have a few more things I'd like to add further down the track, but this is a pretty good starting point.</p><p><strong>Features</strong></p><p><strong>Supports both ref2va and fl2va</strong></p><p>Select a fl2va model and toggle the 'use fl2va?' option to run in fl2va mode, leave it off and load a ref2va model for ref2va mode.</p><p><strong>Automatically constructs prompts</strong></p><p>Each section of a H3 prompt (subject_descriptions, retention_analysis, etc.) has it's own input, all of which are combined to construct the prompt, making it easier to keep track of everything.</p><p><strong>Dynamic prompts</strong></p><p>Uses the <a target="_blank" rel="ugc" href="https://github.com/exectails/comfyui-et_dynamicprompts">dynamic prompts extension</a> to do some pretty neat things, the most useful in this case is using variables in your prompt.</p><p><strong>LoRA Support</strong></p><p>Supports any LoRAs you want. Note: Adding LoRAs will increase generation time.</p><p><strong>Easy references</strong></p><p>Supports the maximum amount of references ref2va uses (9 image, 3 video and 3 audio), and will automatically disable image references 9, 8 and 7 for each reference over the maximum of 12.</p><p><strong>Resize references</strong></p><p>Allows the resizing of image and video references to match the selected generation size, useful for fl2va and keyframes.</p><p><strong>Required node packs</strong></p><p>Uses the following node packs:</p><pre><code>🟢 Will be auto-detected by comfyui-manager
🟡 Can be installed using comfyui-manager, but is not auto-detected (you will have to search for it). Manual installation may be easier to avoid getting the wrong pack.
🔴 Requires manual installation</code></pre><p>🟢 <a target="_blank" rel="ugc" href="https://github.com/ltdrdata/ComfyUI-Impact-Pack">Impact Pack</a></p><p>🟢 <a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-KJNodes">KJ Nodes</a></p><p>🟢 <a target="_blank" rel="ugc" href="https://github.com/rgthree/rgthree-comfy">RGThree Comfy</a></p><p>🟡 <a target="_blank" rel="ugc" href="https://github.com/exectails/comfyui-et_dynamicprompts">Dynamic Prompts</a></p><p>🔴 <a target="_blank" rel="ugc" href="https://github.com/T8mars/comfyui-minimax-h3-audio-T8">MiniMax H3 Audio T8</a> Note: This one can be installed using comfyui-manager (it is not auto-detected), but reports it's compatibility requirements incorrectly so it cannot easily be enabled. Manual installation bypasses that.</p><p>🔴 <a target="_blank" rel="ugc" href="https://github.com/yichengup/ComfyUI-YCNodes-MiniMax-H3">YC Nodes MiniMax H3</a></p><p>🔴 <a target="_blank" rel="ugc" href="https://github.com/slmonker/selflift-Avatar">Selflift Avatar</a></p><p><strong>Required models</strong></p><p><strong>Diffusion model</strong></p><p>Any int8-convrot model, I use <a target="_blank" rel="ugc" href="https://huggingface.co/smhfacct/Minimax-H3-fl2va-ref2va-hybrid-models">these hybrid models</a> for ref2va, and the <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors">basic fl2va one</a>. These go in models/diffusion_models</p><p>Text encoder</p><p><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors">This one</a>. Goes in models/text_encoders</p><p><strong>Vae</strong></p><p><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_video_vae_int8_convrot.safetensors">Video</a>, <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/blob/main/vae/minimax_h3_audio_vae_fp32.safetensors">audio</a>. Goes in models/vae</p><p><strong>Latent upscaler</strong></p><p><a target="_blank" rel="ugc" href="https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler/blob/main/minimax_h3_latent_upscaler_3d_conv_v1/minimax_h3_latent_upscaler_3d_conv_v1_bf16.safetensors">This one</a>. Goes in models/latent_upscale_models</p><p><strong>4 Step LoRA</strong></p><p>Any pruned LoRA model (must be pruned, will crash otherwise). I use <a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MiniMax-H3-experimental/blob/main/loras/minimax_h3_4step_lora_flashgen_v1.0_768p_fl2va_pruned_avg_rank_13_bf16.safetensors">this one</a>, <a target="_blank" rel="ugc" href="https://civitai.red/models/2837571/minimax-h3-turbo-loras?modelVersionId=3202732">this one</a> is also pretty good. Goes in models/loras</p><p><strong>Full model paths</strong></p><pre><code>📂 ComfyUI
├── 📂 models
|   ├── 📂 diffusion_models
|   |    ├── DIFFUSION MODELS GO HERE
|   |
|   ├── 📂 latent_upscale_models
|   |    ├── LATENT UPSCALER GOES HERE
|   |
|   ├── 📂 loras
|   |    ├── 4 STEP LORA GOES HERE
|   |
|   ├── 📂 text_encoders
|   |    ├── TEXT ENCODER GOES HERE
|   |
|   ├── 📂 vae
|   |    ├── VAE MODELS GO HERE</code></pre>

Files with SHA256: 892d6fe3e9fad7f50cbf817124fa394b5ce7e19f4e55905423522038053bc51e