---
title: "MATLOWAI Rebuilds MiniMax H3 to Run Video and Audio on 16GB Cards"
subtitle: "A 21 GB single-file build of MiniMax-H3 renders 10-second clips with synchronized stereo audio in about 76 seconds at 4 steps."
url: "https://alphasignal.ai/news/matlowai-rebuilds-minimax-h3-to-run-video-and-audio-on-16gb-cards"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-08-29T22:04:12.000Z"
updated: "2026-09-05T07:36:29.841Z"
type: "Model"
topics:
  - "Video"
  - "Gpus"
subtopics:
  - "Video Generation"
  - "Quantization"
upvotes: 35
source_url: "https://huggingface.co/MATLOWAI/minimax-h3-fused-turbo-int8-convrot"
comments: 0
truncated: true
---

# MATLOWAI Rebuilds MiniMax H3 to Run Video and Audio on 16GB Cards

> A 21 GB single-file build of MiniMax-H3 renders 10-second clips with synchronized stereo audio in about 76 seconds at 4 steps.

A community rebuild of MiniMax-H3 is quietly turning the open-weights video model into something you can actually run on a workstation. The [MATLOWAI/minimax-h3-fused-turbo-int8-convrot](https://huggingface.co/MATLOWAI/minimax-h3-fused-turbo-int8-convrot) release packs the base transformer, a distilled turbo LoRA, and a motion-smoothing LoRA into one INT8 file that generates video and native stereo audio in the same pass, and it has already pulled nearly 20,000 downloads.

[MiniMax H3](https://blog.comfy.org/p/minimax-h3-day-0-support-in-comfyui) accepts text, images, video, or audio and outputs video with real stereo sound, up to 2K resolution and 15 seconds per clip. Voice, sound effects, and music are modeled together in a single forward pass rather than layered on afterward. This community build is a pragmatic engineering pass on top of that base, aimed at making iteration cheap.

## What's actually in the file

The author bakes three things together offline, then quantizes once. The base is the pruned FL2VA transformer with a rank-1024 SVD of the reference-to-video weight delta fused in, so a single partition handles first/last-frame conditioning and reference-to-video without checkpoint swapping. On top of that, lightx2v's 8-step turbo LoRA is merged at strength 1.0 and the Mystic v2.0 motion-smoothing LoRA at 0.7.

The ConvRot in the name comes from the quantization step. In each of the 50 transformer blocks, the four heavy Linear weights (`qkv_proj`, `out_proj`, `fc1`, `fc2`) are cast to INT8 with per-channel ConvRot at group 256, using ComfyUI's native

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [MATLOWAI/minimax-h3-fused-turbo-int8-convrot](https://huggingface.co/MATLOWAI/minimax-h3-fused-turbo-int8-convrot)
- [MiniMax H3 Day-0 Support in ComfyUI](https://blog.comfy.org/p/minimax-h3-day-0-support-in-comfyui)
- [MiniMax H3 in ComfyUI: T2V, I2V, and R2V Video Workflows](https://docs.comfy.org/tutorials/video/minimax/minimax-h3)
- [Awesome MiniMax-H3](https://github.com/wildminder/awesome-minimax-H3)
- [MiniMax H3 on Comfy: Open-Weight Video Model](https://comfy.org/minimax-h3/)