GE

gemma 4 12B it heretic QAT

GemmaGGUF by SC117Updated 2026-10-03
Download on Hugging Face
Parameters11.9B
Context256K
Layers48
Architecturegemma4
Smallest quant6.7 GB
Quick answer

gemma 4 12B it heretic QAT needs about 20.1 GB of memory at Q4_K_XL (a 6.7 GB file) with an 8K context window, so it fits on a 24 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 51.5 GB.

Gemma 4 12B IT Heretic QAT – 4‑bit GGUF for multimodal inference

Gemma 4 12B IT Heretic QAT is a 12‑billion parameter dense model quantized to Q4_K_XL. It supports text and image input with a 262,144‑token context window. The model is released under Apache‑2.0 and is ready for deployment via llama.cpp.

Good at

  • Multimodal text‑image generation
  • Coding assistance
  • Reasoning tasks

Limits

  • Requires ~20 GB VRAM at 8 k context
  • No vision support in some runtimes
  • License is Apache‑2.0

Best for

  • Chatbots needing image input
  • Code generation with image context
  • Long‑form reasoning with large context

Who it’s for. Developers building multimodal applications on laptops or servers. Researchers wanting a 12B dense model with near‑bfloat16 quality in a 4‑bit format.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_XL6.7 GB 20.1 GB
Too big Get file →

The Q4_K_XL file is 6.72 GB and needs about 20 GB VRAM at 8 k context. For 8 GB VRAM use a lower‑precision quant like Q4_0 if available. For 12–16 GB VRAM the Q4_K_XL is suitable.

What GPU do you need for gemma 4 12B it heretic QAT?

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: no quant fits at an 8K context.
  • 24 GB VRAM: the best quant that fits is Q4_K_XL (6.7 GB file, about 20.1 GB in use).

Which gemma 4 12B it heretic QAT quant should you choose?

Q4_K_XL (6.7 GB) is the middle option in this repository. 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

Architecturegemma4
Parameters11.9B
Layers48
Trained context262,144 tokens
KV cache (fp16)1536 KB per token

Source

Quantized bySC117
FamilyGemma
Licenseapache-2.0
Downloads (HF)12,873
Last updated2026-10-03

Frequently asked questions about gemma 4 12B it heretic QAT

How much VRAM does gemma 4 12B it heretic QAT need?

About 20.1 GB at Q4_K_XL with an 8K context window. That is the 6.7 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 gemma 4 12B it heretic QAT on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 20.1 GB, so you would need a larger GPU or CPU offloading.

What is the file size of gemma 4 12B it heretic QAT GGUF?

It depends on the quant: Q4_K_XL is 6.7 GB.

What context length does gemma 4 12B it heretic QAT support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 12.9 GB at 8K and 51.5 GB at 32K.

Where can I download gemma 4 12B it heretic QAT GGUF?

The files are on Hugging Face in the SC117/gemma-4-12B-it-heretic-QAT-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 gemma 4 12B it heretic QAT 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?

No, the Q4_K_XL variant requires approximately 20 GB VRAM at 8 k context. Use a lower‑precision quant if available.

Does this model support vision?

Yes, it handles image input as part of its multimodal capabilities.

What license applies?

The model is released under the Apache‑2.0 license.

Other GGUFs of gemma-4-12B-it-qat-q4_0-unquantized-heretic

Similar models

Browse all →

More from SC117

All by SC117 →

More Gemma models

All Gemma →

Other medium models

All medium models →