comfy-MiniMax-H3-workflows

MiniMax-H3 INT8 FLF, R2V, Bridge & TTS Workflows Collection

Optimized ComfyUI workflows for MiniMax-H3, featuring INT8 inference, native Block Sparse Attention, Spectrum, Lightx2v LoRA, Turbo-LoRA, native Extend, Fun ControlNet Union V2V control, Face Refine, TTS, Latent Upscaler, Bridge, and Group Bypass.

Workflows

Base model:

Comfy-Org/MiniMax-H3
https://huggingface.co/Comfy-Org/MiniMax-H3

Features

INT8

• Reduced VRAM usage
• Faster inference
• Improved memory efficiency

Block Sparse Attention

The workflows use the native Block Sparse Attention node for optimized MiniMax-H3 attention processing.

Spectrum

ComfyUI-Spectrum-MiniMax-H3
https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3

Optimized sampling and MiniMax-H3 inference acceleration.

Lightx2v LoRA

lightx2v/Minimax-h3-Turbo
https://huggingface.co/lightx2v/Minimax-h3-Turbo

Acceleration LoRA for reduced sampling steps.

Turbo-LoRA

MiniMax-H3-Turbo-Lora-ComfyUI
https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI

Experimental low-step acceleration.

Native Extend

MiniMax-H3 supports native video extension for continuing existing video content.

Fun ControlNet Union

Integrated into the I2V and R2V workflows for V2V control.

Supported controls:

• OpenPose
• Canny
• Depth

Model:

Kijai/MiniMax-H3-experimental
https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main/controlnet

Available:

• minimax_h3_fun_controlnet_union_pruned_bf16.safetensors
• minimax_h3_fun_controlnet_union_pruned_int8_convrot.safetensors

Latent Upscaler

ComfyUI MiniMax-H3 Latent Upscaler
https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler

Integrated into I2V and R2V.

Low-resolution generation → Latent Upscaling → High-resolution decoding

Bridge

Generates an intermediate video between two existing clips:

Video A → Generated Bridge → Video B

Group Bypass

ComfyUI-Switchboard
https://github.com/Oratorian/ComfyUI-Switchboard

Used to enable or bypass optional workflow groups.

R2V Reference LoRA

Required for R2V:

minimax_h3_ref_lora_rank_256_bf16.safetensors

https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main/loras

Face Refine

ComfyUI-H3-FaceRefine
https://github.com/Carasibana/ComfyUI-H3-FaceRefine

Provides automatic face detection and refinement.

Text-to-Speech

MiniMax-H3 can generate audio only, bypassing video generation.

Workflow Details

First-Last-Frame (FLF)

MiniMax_int8_I2V-javanoYYMM.x.json

• First + Last Frame conditioning
• Fun ControlNet Union / V2V
• OpenPose / Canny / Depth control
• Native Extend
• Latent Upscaler
• Native Block Sparse Attention
• INT8
• Spectrum
• Lightx2v LoRA
• Optional Turbo-LoRA
• Group Bypass

Reference-to-Video (R2V)

MiniMax_int8_R2V-javanoYYMM.x.json

• Reference image conditioning
• Reference LoRA
• Fun ControlNet Union / V2V
• OpenPose / Canny / Depth control
• Native Extend
• Latent Upscaler
• Native Block Sparse Attention
• INT8
• Spectrum
• Lightx2v LoRA
• Optional Turbo-LoRA
• Group Bypass

Required LoRA:

minimax_h3_ref_lora_rank_256_bf16.safetensors

Bridge

MiniMax_int8_Bridge-javanoYYMM.x.json

Video A → Generated Bridge → Video B

Creates a generated transition between two existing video clips.

Face Refine

MiniMax_int8_FR-javanoYYMM.x.json

Uses ComfyUI-H3-FaceRefine for automatic face refinement.

Text-to-Speech

MiniMax_int8_TTS-javanoYYMM.x.json

Audio-only MiniMax-H3 TTS workflow.

Required Custom Nodes

• ComfyUI-Spectrum-MiniMax-H3
  https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3
• ComfyUI-H3-FaceRefine
  https://github.com/Carasibana/ComfyUI-H3-FaceRefine
• ComfyUI MiniMax-H3 Latent Upscaler
  https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler
• ComfyUI-Switchboard
  https://github.com/Oratorian/ComfyUI-Switchboard

Block Sparse Attention is provided by the native ComfyUI node and requires no additional custom node.

Required Models

Main Model

Comfy-Org/MiniMax-H3
https://huggingface.co/Comfy-Org/MiniMax-H3

R2V Reference LoRA

minimax_h3_ref_lora_rank_256_bf16.safetensors

https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main/loras

Fun ControlNet Union

Kijai/MiniMax-H3-experimental
https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main/controlnet

• minimax_h3_fun_controlnet_union_pruned_bf16.safetensors
• minimax_h3_fun_controlnet_union_pruned_int8_convrot.safetensors

Lightx2v LoRA

lightx2v/Minimax-h3-Turbo
https://huggingface.co/lightx2v/Minimax-h3-Turbo

Optional Turbo-LoRA

MiniMax-H3-Turbo-Lora-ComfyUI
https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI

Latent Upscaler

Minimax_h3_latent_Upscaler
https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler

Installation

1. Update ComfyUI and ComfyUI-Manager.
2. Install the required custom nodes.
3. Download the MiniMax-H3 model.
4. Download the required R2V Reference LoRA if using R2V.
5. Download a Fun ControlNet Union model if using V2V.
6. Download Lightx2v LoRA and/or Turbo-LoRA if required.
7. Download the Latent Upscaler model if using Latent Upscaler.
8. Load the desired workflow.

Workflow files:

• MiniMax_int8_I2V-javanoYYMM.x.json
• MiniMax_int8_R2V-javanoYYMM.x.json
• MiniMax_int8_Bridge-javanoYYMM.x.json
• MiniMax_int8_FR-javanoYYMM.x.json
• MiniMax_int8_TTS-javanoYYMM.x.json

Notes

• FLF = First-Last-Frame.
• R2V = Reference-to-Video.
• V2V = Video-to-Video.
• FR = Face Refine.
• TTS = Text-to-Speech.
• Extend is natively supported by MiniMax-H3.
• Native Block Sparse Attention is integrated into the I2V and R2V workflows.
• Fun ControlNet Union is integrated into the I2V and R2V workflows.
• The R2V workflow requires minimax_h3_ref_lora_rank_256_bf16.safetensors.
• V2V requires a compatible Fun ControlNet Union model.
• TTS outputs audio only.
• Latent Upscaling uses low-resolution generation followed by latent upscaling.
• ComfyUI-Switchboard controls Group Bypass.
• Turbo-LoRA is experimental.
• Combining multiple acceleration methods may require tuning sampling steps and LoRA weights.
• If audio is not generated correctly, try lowering shift_audio in the ModelSamplingMiniMaxH3 node inside the INT8 Model Loader subgraph from 4.00 to 3.00.

Motion Context Compatibility Note

The ComfyUI-H3-Motion-Context custom node is no longer required by the latest workflows.

Starting with:

• I2V: 2608.9
• R2V: 2608.16

Motion Context has been replaced with the native MiniMaxH3AddGuide node.

Additional Resources

MiniMax-H3
https://huggingface.co/Comfy-Org/MiniMax-H3

R2V Reference LoRA
https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main/loras

Fun ControlNet Union
https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main/controlnet

Spectrum MiniMax-H3
https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3

H3 Face Refine
https://github.com/Carasibana/ComfyUI-H3-FaceRefine

Lightx2v LoRA
https://huggingface.co/lightx2v/Minimax-h3-Turbo

MiniMax-H3-Turbo-Lora-ComfyUI
https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI

MiniMax-H3 Latent Upscaler
https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler

MiniMax-H3 Latent Upscaler Model
https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler

ComfyUI-Switchboard
https://github.com/Oratorian/ComfyUI-Switchboard

License

Please follow the respective licenses of MiniMax-H3, ComfyUI, the listed custom nodes, LoRAs, Fun ControlNet Union models, the Latent Upscaler, and other third-party components used by these workflows.