ComfyUI model file · Diffusion model (fine-tune)

Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors

Minimax-h3 Singularity Ref2VA v1.3 (pruned int8) · Community fine-tune of MiniMax H3 Ref2VA

This is Singularity v1.3, a community fine-tune of MiniMax H3’s Ref2VA model by WarmBloodAban, pruned and stored in int8. It drops in where the official Ref2VA model goes: ComfyUI/models/diffusion_models, loaded with Load Diffusion Model.

  • MiniMax H3 fine-tune
  • Ref2VA
  • 20.97 GB
  • int8

The author spells it Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors. Workflows often use the lowercase name; ComfyUI on Windows and macOS treats both as the same file, Linux does not.

Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors

Type
Diffusion model (fine-tune)
Format
Safetensors · int8
Size
20.97 GB (20,967,647,456 bytes)
SHA256
412a7b12…622949cce
ComfyUI folder
models/diffusion_models/
SHA256 and size from Hugging Face file metadata (WarmBloodAban/Minimax-h3_Singularity), matching the Civitai API hash, checked by RunComfy on .

The short version

What is Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors?

Singularity is a fine-tune of the MiniMax H3 Ref2VA model published by WarmBloodAban. It keeps H3’s reference-to-video-and-audio behaviour and the same loaders, so any Ref2VA workflow can switch to it by changing the file in Load Diffusion Model.

This file is v1.3, pruned and int8. The same repo also has a full v1.3 int8 (34.0 GB), a pruned w4a8 (11.8 GB) and the newer v2 builds.

Install

Download and install Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors

Where it goes
ComfyUI/
└─ models/
   ├─ …
   └─ diffusion_models/
      └─ Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors
terminal · Linux / macOS
$ cd ComfyUI/models/diffusion_models$ wget -O Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors \$   https://huggingface.co/WarmBloodAban/Minimax-h3_Singularity/resolve/main/Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors$ sha256sum Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors# 412a7b126595a958964193f3b42513d7cf2df4196ef04223a0f05e3622949cce

Safetensors files hold tensors only and cannot run code. Still compare the SHA256: a failed download often leaves an HTML error page saved under the model’s filename.

Use

How to load Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors in ComfyUI

Load Diffusion Model reads the diffusion model and outputs a MODEL for the sampler. The text encoder and VAE load in their own nodes.

Minimal graph. Click a node to open its reference page.

Works with

Files a MiniMax H3 video workflow needs

Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors is one part of the stack. A workflow that reports a missing model usually needs these too. Use the Ref2VA diffusion model instead of FL2VA for reference-to-video workflows.

FileRoleFolderSizeDownload
qwen3vl_32b_minimax_h3_nvfp4_awq.safetensorsText encoder (Qwen3-VL 32B)models/text_encoders/15.69 GB Get
minimax_h3_video_vae_fp16.safetensorsVideo VAEmodels/vae/5.21 GB Get
minimax_h3_audio_vae_fp32.safetensorsAudio VAEmodels/vae/605 MB Get
minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensorsTurbo LoRA, 4 steps(optional)models/loras/1.96 GB Get

Run it

ComfyUI workflows that can use Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors

We match workflows by the model files they load: this exact file first, then other versions of it, then workflows built for the same use, then other models from the same family.

  1. Uses this exact file1
  2. Built for reference-to-video25

Uses this exact file

1 workflow

These workflows already load Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors. Open one and run it as is.

Built for reference-to-video

25 workflows

These workflows load another reference-to-video model in the same folder, or they are reference-to-video workflows that already have a Load Diffusion Model node. Select Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors there.

Compare

Same file, other names, and alternatives

FileRelationSizeBest for
Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensorsThis page · v1.3 pruned int820.97 GBDrop-in for pruned Ref2VA workflows
Minimax-h3_Singularity_ref2va_v1.3_Pruned_w4a8.safetensorsSame version · w4a811.77 GBLess VRAM
Minimax-h3_Singularity_ref2va_v1.3_int8.safetensorsSame version · full (not pruned) int834.00 GBWorkflows built for the full model
Minimax-h3_Singularity_ref2va_v2_pruned_int8.safetensorsNewer version · v2 pruned int820.97 GBLatest Singularity release
minimax_h3_ref2va_pruned_fp8_scaled.safetensorsOfficial base model20.96 GBStock MiniMax H3 behaviour

Troubleshoot

Common errors and fixes

Value not in list: unet_name: 'Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors' not in [...]

Why: The workflow asks for this exact filename, but ComfyUI cannot find it in models/diffusion_models. Usually the file is missing, sits in a different folder, or was saved under another name such as minimax-h3_singularity_ref2va_pruned_v1.3_int8.safetensors.

Fix: Move the file into ComfyUI/models/diffusion_models/, rename it to match the workflow (or pick it again in the dropdown), then refresh the node list or restart ComfyUI.

SafetensorError: Error while deserializing header

Why: The download is incomplete, or the browser saved an HTML sign-in or error page under the model’s filename.

Fix: Check that the file is exactly 20,967,647,456 bytes and that its SHA256 matches the value above. Download it again from Hugging Face · WarmBloodAban/Minimax-h3_Singularity if not.

Rather skip the setup?

Run MiniMax H3 Multi-Character Swap in your browser on a cloud GPU. Nothing to download.

Generate video with MiniMax H3 online

FAQ

Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors FAQ

Does the capitalization of the filename matter?

Only on Linux, where ComfyUI treats Minimax-h3_Singularity… and minimax-h3_singularity… as different files. If a workflow cannot find it, rename the file to match the workflow exactly or pick it again in Load Diffusion Model.

Where do I put Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors in ComfyUI?

In ComfyUI/models/diffusion_models/. Then refresh the node list or restart ComfyUI so Load Diffusion Model lists the file.

Which ComfyUI node loads Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors?

Load Diffusion Model. It is built into ComfyUI, so no custom nodes are needed.

Can I use Minimax-h3_Singularity_ref2va_Pruned_v1.3_int8.safetensors commercially?

The repo is tagged Apache-2.0, but this is a fine-tune of MiniMax H3, so the MiniMax H3 Community License still applies to the base weights: show “MiniMax H3” in commercial products, and get MiniMax’s permission above US$20 million yearly revenue.

Sources: WarmBloodAban/Minimax-h3_Singularity · Civitai model 2917208 · MiniMaxAI/MiniMax-H3

RunComfy
Copyright 2026 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.