Gemma4 12B IT Abliterated needs about 20.8 GB of memory at Q4_K_M (a 7.4 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 7 quants, from 6.6 GB (Q3_K_L) to 12.7 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 51.5 GB.
Gemma4 12B IT Abliterated – a 12B Gemma model with refusal direction removed
Gemma4 12B IT Abliterated is a 12‑billion‑parameter decoder‑only model with 48 layers and 3840 hidden dimension. It supports 128K token context and multimodal inputs (vision and audio) via a single linear projection. The model is released under Apache‑2.0.
Good at
- Large context reasoning
- Vision and audio multimodal tasks
- Instruction following without base‑model refusals
Limits
- No separate encoder towers for vision/audio
- Requires large VRAM for full‑precision inference
- No official benchmarks beyond A100‑40GB
Best for
- Research on refusal removal
- Red‑team testing of LLM safety
- Creative applications needing unrestricted instruction compliance
Who it’s for. Researchers, security analysts, and developers who need a Gemma‑4‑12B‑IT variant that complies with all prompts. The model is suitable for experimentation and red‑team scenarios where base‑model refusals are obstacles.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_K_L | 6.6 GB | 20.0 GB | Too big | Get file → | |
| Q4_K_M | 7.4 GB | 20.8 GB | Too big | Get file → | |
| Q4_K_M | 7.4 GB | 20.8 GB | Too big | Get file → | |
| Q5_K_M | 8.5 GB | 21.9 GB | Too big | Get file → | |
| Q5_K_M | 8.5 GB | 21.9 GB | Too big | Get file → | |
| Q8_0 | 12.7 GB | 26.1 GB | Too big | Get file → | |
| Q8_0 | 12.7 GB | 26.1 GB | Too big | Get file → |
For 8 GB VRAM, use Q3_K_L (6.57 GB) or Q4_K_M (7.38 GB) – memory usage is about 20 GB at 8 k context. For 12 GB VRAM, Q5_K_M (8.55 GB) is recommended. Q8_0 (12.67 GB) requires 26 GB VRAM and is best for maximum fidelity.
What GPU do you need for Gemma4 12B IT Abliterated?
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 Q5_K_M (8.5 GB file, about 21.9 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (12.7 GB file, about 26.1 GB in use).
Which Gemma4 12B IT Abliterated quant should you choose?
Q4_K_M is the usual starting point for Gemma4 12B IT Abliterated: it balances file size and quality well. Here it is a 7.4 GB file.
The Q8_0 file is 12.7 GB, about 1.7 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.
If memory is tight, the smallest file is Q3_K_L at 6.6 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Gemma4 12B IT Abliterated
How much VRAM does Gemma4 12B IT Abliterated need?
About 20.8 GB at Q4_K_M with an 8K context window. That is the 7.4 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 Gemma4 12B IT Abliterated on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 20.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Gemma4 12B IT Abliterated GGUF?
It depends on the quant: Q3_K_L is 6.6 GB, Q4_K_M is 7.4 GB, Q4_K_M is 7.4 GB, Q5_K_M is 8.5 GB, Q5_K_M is 8.5 GB, Q8_0 is 12.7 GB and 1 more.
What context length does Gemma4 12B IT Abliterated support?
The GGUF header lists a trained context length of 131,072 tokens (128K). 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 Gemma4 12B IT Abliterated GGUF?
The files are on Hugging Face in the DuoNeural/Gemma4-12B-IT-Abliterated-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 Gemma4 12B IT Abliterated 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 Gemma4 12B IT Abliterated on 8 GB VRAM?
Use Q3_K_L or Q4_K_M quantizations; they fit within 8 GB VRAM but will use ~20 GB of VRAM at 8 k context.
Does this model support vision input?
Yes, it has an encoder‑free vision pathway that projects 48×48‑pixel patches directly into the hidden space.
What license applies to this model?
The model is released under the Apache‑2.0 license.