k 12b needs about 11.0 GB of memory at Q4_K_M (a 7.3 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 11 quants, from 4.8 GB (Q2_K) to 12.5 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 12.9 GB.
k 12b – a 11.8B Gemma3 model fine‑tuned for vision tasks
k 12b is a 11.8B parameter Gemma3 model fine‑tuned from unsloth/gemma-3-12b-it. It supports vision and is released under apache-2.0. The model was trained with Unsloth and Huggingface TRL, offering a compact GGUF format.
Good at
- Vision inference
- Text generation
- Fine‑tuned Gemma3 architecture
Limits
- No text‑only context beyond 131072 tokens
- Requires at least 8.5 GB VRAM for Q2_K
- License is apache‑2.0
Best for
- Image captioning
- Vision‑enabled chat
- Fine‑tuned inference on GPUs
Who it’s for. Developers needing a vision‑capable Gemma3 model with a small memory footprint. Ideal for edge deployments where VRAM is limited but a 12 B parameter model is required.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 4.8 GB | 8.5 GB | Too big | Get file → | |
| Q3_K_S | 5.5 GB | 9.2 GB | Too big | Get file → | |
| Q3_K_M | 6.0 GB | 9.7 GB | Too big | Get file → | |
| Q3_K_L | 6.5 GB | 10.2 GB | Too big | Get file → | |
| IQ4_XS | 6.6 GB | 10.3 GB | Too big | Get file → | |
| Q4_K_S | 6.9 GB | 10.7 GB | Too big | Get file → | |
| Q4_K_M | 7.3 GB | 11.0 GB | Too big | Get file → | |
| Q5_K_S | 8.2 GB | 12.0 GB | Too big | Get file → | |
| Q5_K_M | 8.4 GB | 12.2 GB | Too big | Get file → | |
| Q6_K | 9.7 GB | 13.4 GB | Too big | Get file → | |
| Q8_0 | 12.5 GB | 16.2 GB | Too big | Get file → |
For 8 GB VRAM, use Q2_K (4.77 GB file, 8.5 GB memory). If 10 GB VRAM is available, Q4_K_S (6.94 GB file, 10.7 GB memory) is a good balance. For 16 GB VRAM, Q8_0 (12.51 GB file, 16.2 GB memory) provides the highest quality.
What GPU do you need for k 12b?
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: the best quant that fits is Q5_K_S (8.2 GB file, about 12.0 GB in use).
- 16 GB VRAM: the best quant that fits is Q6_K (9.7 GB file, about 13.4 GB in use).
- 24 GB VRAM: the best quant that fits is Q8_0 (12.5 GB file, about 16.2 GB in use).
Which k 12b quant should you choose?
Q4_K_M is the usual starting point for k 12b: it balances file size and quality well. Here it is a 7.3 GB file.
The Q8_0 file is 12.5 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 Q2_K at 4.8 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about k 12b
How much VRAM does k 12b need?
About 11.0 GB at Q4_K_M with an 8K context window. That is the 7.3 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 k 12b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 8.5 GB, so you would need a larger GPU or CPU offloading.
What is the file size of k 12b GGUF?
It depends on the quant: Q2_K is 4.8 GB, Q3_K_S is 5.5 GB, Q3_K_M is 6.0 GB, Q3_K_L is 6.5 GB, IQ4_XS is 6.6 GB, Q4_K_S is 6.9 GB and 5 more.
What context length does k 12b support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 3.2 GB at 8K and 12.9 GB at 32K.
Where can I download k 12b GGUF?
The files are on Hugging Face in the mradermacher/k-12b-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 k 12b 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 k 12b on 8 GB VRAM?
Yes, using the Q2_K quant which requires 8.5 GB memory at 8k context. It may be close to the limit, so a slightly larger VRAM is recommended.
Does k 12b support text‑only inference?
The model has vision capabilities but can also perform text generation. However, its context is limited to 131072 tokens.
What license applies to k 12b?
The model is released under the apache-2.0 license, allowing commercial use with attribution.