gemma 4 26B A4B it qat heretic needs about 22.8 GB of memory at Q4_K_XL (a 14.2 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 2 quants, from 14.2 GB (Q4_K_XL) to 50.5 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 32.2 GB.
Gemma 4 26B A4B Heretic QAT – 26 B parameters, 4 B active MoE, 256K context, 14 GB Q4_0 GGUF
Gemma 4 26B A4B Heretic QAT is a 26 B parameter multimodal model with 4 B active experts. It supports text and image input and a 256 K token context window. The QAT‑optimized Q4_0 GGUF file is 14 GB and requires about 23 GB VRAM at 8 K context.
Good at
- Multimodal text‑image generation
- Large‑context reasoning
- Coding tasks
Limits
- Requires 8 K context for VRAM estimate
- No audio support in this variant
- Apache 2.0 license limits commercial use without attribution
Best for
- Chatbots needing long context
- Image‑guided code generation
- Research on MoE architectures
Who it’s for. Developers and researchers who need a large multimodal model with a 256 K token window and can run the 14 GB Q4_0 GGUF on GPUs with sufficient VRAM. The model is suitable for servers or workstations with at least 24 GB VRAM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_XL | 14.2 GB | 22.8 GB | Too big | Get file → | |
| F16 | 50.5 GB | 59.1 GB | Too big | Get file → |
For 24–30 GB VRAM, use the Q4_K_XL quantized file (14.25 GB, ~22.8 GB VRAM at 8 K context). If more VRAM is available, the F16 file (50.51 GB, ~59.1 GB VRAM) can be used for higher precision. The estimates are approximate.
What GPU do you need for gemma 4 26B A4B it qat heretic?
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 (14.2 GB file, about 22.8 GB in use).
- 32 GB VRAM: the best quant that fits is Q4_K_XL (14.2 GB file, about 22.8 GB in use).
- 48 GB VRAM: the best quant that fits is Q4_K_XL (14.2 GB file, about 22.8 GB in use).
Which gemma 4 26B A4B it qat heretic quant should you choose?
Q4_K_XL (14.2 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
Frequently asked questions about gemma 4 26B A4B it qat heretic
How much VRAM does gemma 4 26B A4B it qat heretic need?
About 22.8 GB at Q4_K_XL 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 gemma 4 26B A4B it qat heretic on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 22.8 GB, so you would need a larger GPU or CPU offloading.
What is the file size of gemma 4 26B A4B it qat heretic GGUF?
It depends on the quant: Q4_K_XL is 14.2 GB, F16 is 50.5 GB.
What context length does gemma 4 26B A4B it qat heretic support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 8.1 GB at 8K and 32.2 GB at 32K.
Where can I download gemma 4 26B A4B it qat heretic GGUF?
The files are on Hugging Face in the SC117/gemma-4-26B-A4B-it-qat-heretic-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 26B A4B it qat heretic 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 14 GB Q4_0 file requires about 23 GB VRAM at 8 K context. A smaller quantization is not available for this model.
Does it support audio input?
No, audio is supported only in the E2B, E4B, and 12B variants, not in the 26B A4B version.
What license applies?
The model is released under Apache 2.0, allowing free use with attribution and no patent claims.