MiniMax H3 ComfyUI Workflow: Install & Run Guide

Check the model license before downloading

MiniMax H3 community weights have an official license with territory and use conditions. This independent guide is not legal advice. Read the current MiniMax H3 Community License Agreement for your location and use case.

Quickstart

Run MiniMax H3 in five steps

1

Update ComfyUI

Use a recent build, update Manager and restart before importing the workflow.

2

Download files

Choose a supported diffusion model, text encoder, video VAE and optional audio VAE.

3

Use correct folders

Keep diffusion, encoder and VAE files in their separate model directories.

4

Import JSON

Drag the official workflow into ComfyUI and select your installed filenames.

Step 5: queue a short moderate-resolution test. Once it succeeds, save that workflow as your clean baseline before adding cache, LoRAs, longer clips or audio.

Model files

Put each file in the right folder

Use the Comfy-Org MiniMax H3 repository as the current file index. Download only the variants required by your chosen workflow.

ComfyUI/models/diffusion_models/
ComfyUI/models/text_encoders/
ComfyUI/models/vae/

Do not rename files just to match a screenshot. Open the loader dropdown and select the filename that is actually installed. If it is missing, rescan models or restart ComfyUI.

A workflow is a map of dependencies. Resolve red or missing nodes before adjusting generation quality.

Official templates

Choose T2V, I2V or R2V

Workflow	Use it when	Input	Official JSON
T2V	You want the simplest prompt-to-video test	Text prompt	video_minimax_h3_t2v.json
I2V	A source image must control composition, product or identity	Image + prompt	video_minimax_h3_i2v.json
R2V	You need reference media conditioning	Reference media + prompt	video_minimax_h3_r2v.json · setup guide

Import checklist

Verify the workflow before Queue Prompt

Drag the JSON into the canvas or use Open.
Check every loader for an installed diffusion model, text encoder and VAE.
Resolve missing custom nodes through the node's named repository or ComfyUI Manager.
Set the source image or reference input if using I2V or R2V.
Use a deterministic seed while debugging.
Save a clean local copy after the first successful output.

Community workflow

RTX 3060 R2V example

A community author published `MBEDIT-MMH33_r2v_TurboLora_v6.json` for an RTX 3060 12GB, 32GB RAM and Windows 10 setup. The associated report used 1344×768, five seconds, SageAttention, Turbo LoRA, no cache and no audio, with about a ten-minute result.

Open the public workflow directory · Read the original report

Status: Community-reported · Site test pending. Custom workflows can require nodes and model variants beyond the official template.

Common setup errors

Fix the workflow before tuning quality

Model dropdown is empty

Check the folder, filename and file completion, then restart ComfyUI. A partial download may appear on disk but fail to load.

Red missing nodes

Install the exact named custom-node repository. Restart and reload the JSON before installing another pack.

OOM before step one

The text encoder or model handoff may be the problem. A lower sampler resolution cannot fix every pre-sampling OOM.

Output has no audio

Confirm the workflow includes the audio VAE and audio path. Some optimized community workflows intentionally disable audio.

Open the symptom-led troubleshooting guide.

FAQ

MiniMax H3 ComfyUI questions

Where can I download the MiniMax H3 ComfyUI workflow?

Comfy-Org publishes official MiniMax H3 T2V, I2V and R2V JSON templates in its workflow_templates GitHub repository. Choose the task first, download that JSON, and then select the model files installed on your machine.

Where do MiniMax H3 model files go in ComfyUI?

Put diffusion files in ComfyUI/models/diffusion_models, Qwen3-VL text encoders in ComfyUI/models/text_encoders, and video or audio VAE files in ComfyUI/models/vae.

Which MiniMax H3 workflow should a beginner use?

Start with T2V if you only have a prompt. Use I2V when a source image must control composition or identity, and use R2V when the workflow needs reference media. Run a short test before enabling audio or increasing frames.