# minimax_h3_latent_upscaler_3d_fp16.safetensors

> Minimax H3 Latent Upscaler 3D (fp16) · Latent-space upscaler for MiniMax H3 video

This is LBH-123-AI’s 3D latent upscaler for MiniMax H3 in fp16. It enlarges H3 video latents directly, so you can generate at low resolution and refine at high resolution without decoding through the VAE. Put it in ComfyUI/models/latent_upscale_models.

The author later renamed this file minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors. Both names have the same SHA256.

## Facts

- **Same file as:** minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors
- **Type:** Latent upscale model
- **Parameters:** 345M (3D convolution)
- **Format:** Safetensors · fp16
- **Size:** 691 MB (690,592,672 bytes)
- **SHA256:** 043e5a48e161610ef6c3ea974645220354d06fa618abca15f76d084812eb55c2
- **ComfyUI folder:** ComfyUI/models/latent_upscale_models/
- **Loader node:** Minimax H3 Latent Upscaler (3D)
- **License:** Apache-2.0

## What is minimax_h3_latent_upscaler_3d_fp16.safetensors?

MiniMax H3’s video VAE is about 5 billion parameters, so decoding to pixels, upscaling and encoding again is slow. This model upscales the 24-channel H3 latent itself, keeping the time axis, and avoids the ghosting that plain bilinear latent resizing causes.

It was first published as minimax_h3_latent_upscaler_3d_fp16.safetensors, and many workflows still use that name. The author has since moved it to minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors; the SHA256 043e5a48…84812eb55c2 is the same, so either name works. The author also released a lighter 2D model in November 2026.

## Download and install

- [Hugging Face · LBH-123-AI/Minimax_h3_latent_Upscaler](https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler/resolve/main/minimax_h3_latent_upscaler_3d_conv_v1/minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors) · saves as `minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors` · Author’s repo, current filename. Rename it if your workflow asks for the old one.
- [Hugging Face mirror · Konteni2](https://huggingface.co/Konteni2/Minimax_h3_latent_Upscaler/resolve/main/minimax_h3_latent_upscaler_3d_fp16.safetensors) · saves as `minimax_h3_latent_upscaler_3d_fp16.safetensors` · Same SHA256 under the original filename.

```sh
cd ComfyUI/models/latent_upscale_models
wget -O minimax_h3_latent_upscaler_3d_fp16.safetensors \
  https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler/resolve/main/minimax_h3_latent_upscaler_3d_conv_v1/minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors
sha256sum minimax_h3_latent_upscaler_3d_fp16.safetensors
# 043e5a48e161610ef6c3ea974645220354d06fa618abca15f76d084812eb55c2
```

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.

## How to load it in ComfyUI

The upscaler works on latents, between a low-resolution H3 pass and a refining pass at the target size. The Minimax H3 Latent Upscaler (3D) node loads the file and upscales in one step.

- **Why latent upscaling:** It skips the decode → pixel upscale → encode round trip through H3’s large video VAE.
- **Scales:** The 3D model was trained at 2×, 2.25×, 2.5× and 4×; its author reports weaker results below 1.5×.

## Files a MiniMax H3 video workflow needs

Use the Ref2VA diffusion model instead of FL2VA for reference-to-video workflows.

- [`minimax_h3_fl2va_pruned_int8_convrot.safetensors`](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors) · Diffusion model (FL2VA) · `models/diffusion_models/` · 20.97 GB
- [`qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors`](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors) · Text encoder (Qwen3-VL 32B) · `models/text_encoders/` · 15.69 GB
- [`minimax_h3_video_vae_fp16.safetensors`](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors) · Video VAE · `models/vae/` · 5.21 GB
- [`minimax_h3_audio_vae_fp32.safetensors`](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors) · Audio VAE · `models/vae/` · 605 MB
- [`minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors`](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/loras/minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors) · Turbo LoRA, 4 steps · `models/loras/` · 1.96 GB

## Workflows: Uses this exact file

These workflows already load minimax_h3_latent_upscaler_3d_fp16.safetensors. Open one and run it as is.

- [MiniMax H3 Veda Acceleration | 3x Fast Ref-to-Video](https://www.runcomfy.com/comfyui-workflows/minimax-h3-veda-acceleration)
- [MiniMax H3 Long Video Generation | Motion Context V2](https://www.runcomfy.com/comfyui-workflows/minimax-h3-long-video-in-comfyui)
- [MiniMax H3 Singularity Reference To Video ComfyUI Workflow](https://www.runcomfy.com/comfyui-workflows/minimax-h3-singularity-reference-to-video-comfyui)
- [MiniMax H3 Latent Upscaler | BUNNYH3 Two-Pass Workflow](https://www.runcomfy.com/comfyui-workflows/minimax-h3-latent-upscaler-comfyui-two-pass-3d-video)
- [MiniMax H3 8-Step Acceleration | All-in-One Video Workflow](https://www.runcomfy.com/comfyui-workflows/minimax-h3-8-step-acceleration-all-in-one-comfyui-workflow)

## Workflows: Uses other MiniMax H3 models

These workflows load another MiniMax H3 model. Select minimax_h3_latent_upscaler_3d_fp16.safetensors in their Minimax H3 Latent Upscaler (3D) node to use it instead.

- [MiniMax H3 Action Scenes | Guided Action Video](https://www.runcomfy.com/comfyui-workflows/minimax-h3-action-scenes-reference-video-native-audio)

## Same file, other names, and alternatives

- `minimax_h3_latent_upscaler_3d_fp16.safetensors` · This page · original filename · 691 MB · Workflows that use the old name
- `minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors` · Identical bytes · current filename · 691 MB · Same file
- `minimax_h3_latent_upscaler_3d_conv_v1_bf16.safetensors` · Same model · bf16 · 691 MB · GPUs that prefer bf16
- `minimax_h3_latent_upscaler_2d_conv_v1_fp16.safetensors` · Newer 2D model · 364 MB · Half the size, adjustable sharpness; needs the 2D node
- `h3_upscaler_lms_v0.1.safetensors` · Fine-tune of this model · 691 MB · Extra perceived sharpness

## Common errors and fixes

### Value not in list: model_name: 'minimax_h3_latent_upscaler_3d_fp16.safetensors' not in [...]

Why: The workflow asks for this exact filename, but ComfyUI cannot find it in models/latent_upscale_models. Usually the file is missing, sits in a different folder, or was saved under another name such as minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors.

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

### Missing node type: MinimaxH3LatentUpscaler3D

Why: Minimax H3 Latent Upscaler (3D) is not built into ComfyUI. It comes from the Comfyui_Minimax_h3_latent_Upscaler custom node pack.

Fix: Install Comfyui_Minimax_h3_latent_Upscaler with ComfyUI Manager (or git clone it into ComfyUI/custom_nodes), then 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 690,592,672 bytes and that its SHA256 matches the value above. Download it again from Hugging Face · LBH-123-AI/Minimax_h3_latent_Upscaler if not.

## FAQ

### Is minimax_h3_latent_upscaler_3d_conv_v1_fp16 the same file?

Yes. Both names have the SHA256 `043e5a48e161610ef6c3ea974645220354d06fa618abca15f76d084812eb55c2` and are 690,592,672 bytes. The author renamed it when the 2D model came out.

### Can the 2D node load this file?

No. The 2D node loads 2D checkpoints and the 3D node loads 3D checkpoints; the weights and backbone must match.

### Where do I put minimax_h3_latent_upscaler_3d_fp16.safetensors in ComfyUI?

In `ComfyUI/models/latent_upscale_models/`. Then refresh the node list or restart ComfyUI so Minimax H3 Latent Upscaler (3D) lists the file.

### Which ComfyUI node loads minimax_h3_latent_upscaler_3d_fp16.safetensors?

**Minimax H3 Latent Upscaler (3D)**. It comes from the [Comfyui_Minimax_h3_latent_Upscaler](https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler) custom node pack, so install that first.

### Can I use minimax_h3_latent_upscaler_3d_fp16.safetensors commercially?

The upscaler is released under Apache-2.0, which allows commercial use. The MiniMax H3 base model keeps its own license.

## Sources

- [LBH-123-AI/Minimax_h3_latent_Upscaler](https://huggingface.co/LBH-123-AI/Minimax_h3_latent_Upscaler)
- [ComfyUI node on GitHub](https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler)

Canonical page: https://www.runcomfy.com/comfyui-workflows/model-files/minimax_h3_latent_upscaler_3d_fp16.safetensors
