MI

MiniMax H3

MinimaxGGUF by leejetUpdated 2026-08-30
Download on Hugging Face
Parameters20.1B
Context—
Layers—
Architecture—
Smallest quant6.7 GB
Quick answer

MiniMax H3 needs about 13.3 GB of memory at Q4_K_M (a 11.4 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 7 quants, from 6.7 GB (Q2_K_M) to 18.8 GB (Q4_K_M).

MiniMax H3 – 20.1B parameter multimodal model in GGUF format

MiniMax H3 is a 20.1B parameter multimodal model available as GGUF files. It supports text, image, video, and audio inputs and can generate video with stereo audio up to 2K resolution and 15‑second duration. The model is quantized by leejet and can be used with llama.cpp and ComfyUI.

Good at

  • Multimodal input understanding
  • Video and audio generation
  • High‑resolution 2K output

Limits

  • No vision support in the base GGUF file
  • Large memory requirement for higher quant levels
  • License follows original model

Best for

  • Developers building multimodal generation pipelines
  • Researchers testing multimodal generation
  • Users needing 2K video output

Who it’s for. This model is suitable for developers and researchers who need a general‑purpose multimodal generation system. It is ideal for projects that require text, image, video, and audio processing and generation.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K_M6.7 GB 8.0 GB
Too big Get file →
Q4_K_M11.4 GB 13.3 GB
Too big Get file →
Q4_K_M11.4 GB 13.3 GB
Too big Get file →
Q2_K_M13.1 GB 15.2 GB
Too big Get file →
Q4_K_M18.2 GB 20.9 GB
Too big Get file →
Q4_K_M18.8 GB 21.5 GB
Too big Get file →
Q4_K_M18.8 GB 21.5 GB
Too big Get file →

For 8 GB VRAM use the Q2_K_M 6.72 GB file (est. 8 GB memory at 8 k context). For 13 GB VRAM use the Q4_K_M 11.42 GB file (est. 13.3 GB memory). For 21 GB VRAM use the Q4_K_M 18.78 GB file (est. 21.5 GB memory).

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: the best quant that fits is Q2_K_M (6.7 GB file, about 8.0 GB in use).
  • 12 GB VRAM: the best quant that fits is Q2_K_M (6.7 GB file, about 8.0 GB in use).
  • 16 GB VRAM: the best quant that fits is Q2_K_M (13.1 GB file, about 15.2 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_M (18.8 GB file, about 21.5 GB in use).

Which MiniMax H3 quant should you choose?

Q4_K_M is the usual starting point for MiniMax H3: it balances file size and quality well. Here it is a 11.4 GB file.

If memory is tight, the smallest file is Q2_K_M at 6.7 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Parameters20.1B

Source

Quantized byleejet
FamilyMinimax
Downloads (HF)250,484
Last updated2026-08-30

Frequently asked questions about MiniMax H3

How much VRAM does MiniMax H3 need?

About 13.3 GB at Q4_K_M with an 8K context window. That is the 11.4 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?

Yes, with Q2_K_M: the 6.7 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of MiniMax H3 GGUF?

It depends on the quant: Q2_K_M is 6.7 GB, Q4_K_M is 11.4 GB, Q4_K_M is 11.4 GB, Q2_K_M is 13.1 GB, Q4_K_M is 18.2 GB, Q4_K_M is 18.8 GB and 1 more.

Where can I download MiniMax H3 GGUF?

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

Can I run MiniMax H3 on 8 GB VRAM?

Yes, the Q2_K_M 6.72 GB file requires approximately 8 GB memory at 8 k context.

What is the memory requirement for the Q4_K_M 18.78 GB file?

It requires approximately 21.5 GB memory at 8 k context.

Does the GGUF version support vision?

The provided GGUF files do not include vision support; the model is multimodal but the GGUF quantized files lack vision capability.

Other GGUFs of MiniMax-H3

Similar models

Browse all →

More from leejet

All by leejet →

More Minimax models

All Minimax →

Other medium models

All medium models →