Skip to content
AETRONConstructor
My neuronetsRegistrySpacesDocsFAQ
Sign in
AETRON

A constructor for AI services on AETRON’s decentralised network of GPU miners.

ProductCreate a neuronetModel catalogueDatasetsSpacesMy payouts
ResourcesDocumentationAPITimings & statusesFAQ
Networkaetron.ioPrivacy & verification
© 2026 AETRONTestnet
Models/minimax-h3-gguf-q4-k-m
M

MiniMax H3 (GGUF Q4_K_M, sd.cpp)

MiniMax (GGUF: leejet / Comfy-Org) · MiniMax
Create a neuronet with this model →
VideoMiniMax-H3-CommunitypublicCommunity63B params
text-to-videovideoaudioggufquantizedsd-cpp

Video+stereo audio generation on the sd.cpp engine: T2VA, start/end frame (I2VA/FL2VA). Q4 DiT + Qwen3-VL-32B TE; runs on one 24 GB GPU (needs ~48 GB host RAM).

Video+stereo audio generation on the sd.cpp engine: T2VA, start/end frame (I2VA/FL2VA). Q4 DiT + Qwen3-VL-32B TE; runs on one 24 GB GPU (needs ~48 GB host RAM).

At a glance

  • Task: Video (DiffusionV1)
  • Params: 63B
  • License: MiniMax-H3-Community (commercial use OK)
  • Hardware: 24 GB+ VRAM
  • dtype: Int4
Create a neuronet with this model →Not deploy-ready while hashes are placeholders (revision: PLACEHOLDER_PINNED_SHA).
Overview
Task
Video · DiffusionV1
Family
MiniMax
Tier
Community
Status
active
Runtime spec
dtype
Int4
attention
Sdpa
compile
Eager
seed
0
gpu_arch
Sm89
Runs on
Consumer, Datacenter
Hardware
Min VRAM
24 GB
Recommended
32 GB
Tensor parallel
≥1
TEE archs
Sm90
Verification defaults
Tier
B
Trust
Level1
Security
Standard
Privacy
P0
Min pulse score
95
License
spdx
MiniMax-H3-Community ↗
Commercial use
✓
Redistribution
✓
Source & provenance
Repo
leejet/MiniMax-H3-GGUF ↗, Comfy-Org/MiniMax-H3 ↗
Revision
not pinned
Added
2026-08-21
Verified
2026-08-21