UN

Hy MT2 30B A3B Uncensored v1 APEX

UncensoredGGUF by alphaZimuthUpdated 2026-10-05
Download on Hugging Face
Parameters30B
Context256K
Layers48
Architecturehy_v3
Smallest quant11.6 GB
Quick answer

Hy MT2 30B A3B Uncensored v1 APEX needs about 15.5 GB of memory at its main file (a 14.2 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 3 quants, from 11.6 GB to 19.2 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 3.2 GB.

Hy MT2 30B A3B Uncensored v1 APEX – experimental 30B multilingual model

Hy MT2 30B A3B Uncensored v1 APEX is a 30B parameter model with 3B active parameters in a mixture of experts architecture. It was trained on 262144 context tokens and released under the apache-2.0 license. The model is experimental and may exhibit translation regressions.

Good at

  • Large context handling
  • Multilingual translation
  • Uncensored content generation

Limits

  • Experimental release – limited testing
  • Potential translation regressions
  • No vision support

Best for

  • Research on multilingual translation
  • Testing ablation techniques
  • Comparing quantization variants

Who it’s for. Researchers and developers interested in experimental multilingual models and ablation studies. Use with caution and report issues.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—11.6 GB 12.9 GB
Too big Get file →
—14.2 GB 15.5 GB
Too big Get file →
—19.2 GB 20.5 GB
Too big Get file →

The unknown quant variants range from 11.62 GB to 19.2 GB. For 8 GB VRAM, the 11.62 GB file is the smallest but still exceeds VRAM; consider using a higher VRAM system or a smaller quant if available. Memory estimates are approximate.

What GPU do you need for Hy MT2 30B A3B Uncensored v1 APEX?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) 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 the main file (14.2 GB file, about 15.5 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (19.2 GB file, about 20.5 GB in use).

Which Hy MT2 30B A3B Uncensored v1 APEX quant should you choose?

If memory is tight, the smallest file is the first one in the table at 11.6 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturehy_v3
Parameters30B (approximate, mixture of experts)
Active parameters3B
Layers48
Trained context262,144 tokens
KV cache (fp16)96 KB per token

Source

Quantized byalphaZimuth
Licenseapache-2.0
Downloads (HF)262
Last updated2026-10-05

Frequently asked questions about Hy MT2 30B A3B Uncensored v1 APEX

How much VRAM does Hy MT2 30B A3B Uncensored v1 APEX need?

About 15.5 GB at its main file with an 8K context window. That is the 14.2 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 Hy MT2 30B A3B Uncensored v1 APEX on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 12.9 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Hy MT2 30B A3B Uncensored v1 APEX GGUF?

It depends on the quant: file is 11.6 GB, file is 14.2 GB, file is 19.2 GB.

What context length does Hy MT2 30B A3B Uncensored v1 APEX support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.2 GB at 32K.

Where can I download Hy MT2 30B A3B Uncensored v1 APEX GGUF?

The files are on Hugging Face in the alphaZimuth/Hy-MT2-30B-A3B-Uncensored-v1-APEX-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 Hy MT2 30B A3B Uncensored v1 APEX use?

Hugging Face lists the license as apache-2.0. 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 this model on 8 GB VRAM?

The smallest file is 11.62 GB, which exceeds 8 GB VRAM. It is not recommended to run on 8 GB VRAM without additional memory optimizations.

Is this model suitable for production use?

No. It is an experimental release with limited testing and potential translation regressions.

What license does this model use?

The model is released under the apache-2.0 license.

Similar models

Browse all →

More from alphaZimuth

All by alphaZimuth →

More Uncensored models

All Uncensored →

Other medium models

All medium models →