MI

Sakura MiniMax H3 Uncensored

MinimaxGGUF by webmp3Updated 2026-10-17
Download on Hugging Face
Parameters25.8B
Context—
Layers—
Architecture—
Smallest quant15.1 GB
Quick answer

Sakura MiniMax H3 Uncensored needs about 17.4 GB of memory at Q4_K_M (a 15.1 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 4 quants, from 15.1 GB (Q4_K_M) to 27.5 GB (Q8_0).

25.8B‑parameter text encoder for MiniMax‑H3, uncensored

Sakura MiniMax H3 Uncensored is a 25.8B‑parameter text encoder derived from Qwen3‑VL‑32B‑Instruct. It is released as a GGUF model for use with llama.cpp. The model is uncensored and intended for adult‑only usage. It contains no vision or multimodal components beyond the text encoder.

Good at

  • Large‑scale text generation
  • Uncensored content handling
  • Compatibility with llama.cpp

Limits

  • No vision or audio support
  • Uncensored content requires moderation
  • License is other, not open source

Best for

  • Developers building text‑only pipelines
  • Researchers testing large‑parameter encoders
  • Users needing uncensored text generation

Who it’s for. This model is aimed at developers and researchers who need a large, uncensored text encoder for integration with llama.cpp. It is suitable for applications that require high‑capacity language modeling without multimodal inputs.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M15.1 GB 17.4 GB
Too big Get file →
Q5_K_M18.3 GB 20.9 GB
Too big Get file →
Q6_K21.6 GB 24.7 GB
Too big Get file →
Q8_027.5 GB 31.3 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_M (15.08 GB file, ~17.4 GB VRAM at 8k context). For 12 GB VRAM, Q5_K_M (18.25 GB file, ~20.9 GB VRAM). For 16 GB VRAM, Q6_K (21.62 GB file, ~24.7 GB VRAM). For 32 GB VRAM, Q8_0 (27.49 GB file, ~31.3 GB VRAM). Estimates are approximate.

What GPU do you need for Sakura MiniMax H3 Uncensored?

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: no quant fits at an 8K context.
  • 24 GB VRAM: the best quant that fits is Q5_K_M (18.3 GB file, about 20.9 GB in use).
  • 32 GB VRAM: the best quant that fits is Q8_0 (27.5 GB file, about 31.3 GB in use).

Which Sakura MiniMax H3 Uncensored quant should you choose?

Q4_K_M is the usual starting point for Sakura MiniMax H3 Uncensored: it balances file size and quality well. Here it is a 15.1 GB file.

The Q8_0 file is 27.5 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Parameters25.8B

Source

Quantized bywebmp3
FamilyMinimax
Licenseother
Downloads (HF)46,846
Last updated2026-10-17

Frequently asked questions about Sakura MiniMax H3 Uncensored

How much VRAM does Sakura MiniMax H3 Uncensored need?

About 17.4 GB at Q4_K_M with an 8K context window. That is the 15.1 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 Sakura MiniMax H3 Uncensored on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 17.4 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Sakura MiniMax H3 Uncensored GGUF?

It depends on the quant: Q4_K_M is 15.1 GB, Q5_K_M is 18.3 GB, Q6_K is 21.6 GB, Q8_0 is 27.5 GB.

Where can I download Sakura MiniMax H3 Uncensored GGUF?

The files are on Hugging Face in the webmp3/Sakura-MiniMax-H3-Uncensored-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 Sakura MiniMax H3 Uncensored 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 Sakura MiniMax H3 Uncensored on 8 GB VRAM?

Use the Q4_K_M quantized file, which is 15.08 GB and requires about 17.4 GB VRAM at 8k context. It will not fit on 8 GB VRAM without further quantization.

Does this model support vision or audio?

No. It is a text encoder only and does not include vision or audio components.

What license applies to this model?

The model is released under an "other" license, not an open‑source license.

Other GGUFs of MiniMax-H3

Similar models

Browse all →

More from webmp3

All by webmp3 →

More Minimax models

All Minimax →

Other medium models

All medium models →