MI

MiniMax H3

MinimaxGGUF by DerdisirtlanUpdated 2026-08-14
Download on Hugging Face
Parameters20.1B
Context—
Layers—
Architectureminimax_h3
Smallest quant11.4 GB
Quick answer

MiniMax H3 needs about 17.3 GB of memory at its main file (a 15.0 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 9 quants, from 11.4 GB (Q4_0) to 21.4 GB (Q8_0).

MiniMax H3 – 20.1B omni‑modal generative model

MiniMax H3 is a 20.1B parameter model that supports text, image, video, and audio inputs. It can generate video with native stereo audio up to 2K resolution and 15‑second duration. The model is available in GGUF quantized formats.

Good at

  • Multimodal understanding
  • Text‑to‑video generation
  • Audio‑enabled video output

Limits

  • No vision support in llama.cpp
  • Requires large VRAM for full resolution
  • License type is other

Best for

  • Creating short video clips
  • Generating audio‑video content
  • Multimodal content creation

Who it’s for. Developers and researchers needing a general‑purpose multimodal model for generating short videos and audio‑enabled content. The model is suitable for applications that can handle the required GPU memory and support the GGUF format.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_011.4 GB 13.2 GB
Too big Get file →
Q4_011.4 GB 13.2 GB
Too big Get file →
Q5_013.9 GB 16.1 GB
Too big Get file →
—15.0 GB 17.3 GB
Too big Get file →
—15.0 GB 17.3 GB
Too big Get file →
—20.2 GB 23.1 GB
Too big Get file →
—20.2 GB 23.1 GB
Too big Get file →
Q8_021.4 GB 24.5 GB
Too big Get file →
Q8_021.4 GB 24.5 GB
Too big Get file →

For 8 GB VRAM, use the Q4_0 quantization (11.38 GB file, ~13.2 GB memory at 8k context). For 12 GB VRAM, Q5_0 (13.89 GB file, ~16.1 GB memory). For 24 GB VRAM, Q8_0 (21.41 GB file, ~24.5 GB memory). Estimates are approximate.

What GPU do you need for MiniMax H3?

These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: the best quant that fits is Q4_0 (11.4 GB file, about 13.2 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (20.2 GB file, about 23.1 GB in use).
  • 32 GB VRAM: the best quant that fits is Q8_0 (21.4 GB file, about 24.5 GB in use).

Which MiniMax H3 quant should you choose?

If memory is tight, the smallest file is Q4_0 at 11.4 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureminimax_h3
Parameters20.1B

Source

Quantized byDerdisirtlan
FamilyMinimax
Licenseother
Downloads (HF)1,519
Last updated2026-08-14

Frequently asked questions about MiniMax H3

How much VRAM does MiniMax H3 need?

About 17.3 GB at its main file with an 8K context window. That is the 15.0 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run MiniMax H3 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 13.2 GB, so you would need a larger GPU or CPU offloading.

What is the file size of MiniMax H3 GGUF?

It depends on the quant: Q4_0 is 11.4 GB, Q4_0 is 11.4 GB, Q5_0 is 13.9 GB, file is 15.0 GB, file is 15.0 GB, file is 20.2 GB and 3 more.

Where can I download MiniMax H3 GGUF?

The files are on Hugging Face in the Derdisirtlan/MiniMax-H3-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does MiniMax H3 use?

Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run MiniMax H3 on 8 GB VRAM?

Use the Q4_0 quantized file; it requires about 13.2 GB memory at 8k context, which is close to 8 GB but may need swapping.

Does MiniMax H3 support image inputs?

Yes, it supports up to two input images in FL2VA mode and up to nine images in Ref2VA mode.

What is the maximum video duration MiniMax H3 can generate?

It can generate video up to 15 seconds long at 2K resolution with native stereo audio.

Other GGUFs of MiniMax-H3

Similar models

Browse all →

More Minimax models

All Minimax →

Other medium models

All medium models →