Shisa De 1 needs about 25.3 GB of memory at Q4_K_M (a 16.8 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 11 quants, from 10.6 GB (Q2_K) to 26.9 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 32.2 GB.
Shisa De 1 – a decision‑oriented Gemma4 model with vision and tools support
Shisa De 1 is a 25.2B‑parameter Gemma4 model trained for decision tasks. It outputs typed answers via a restricted option‑letter distribution. The vision encoder is included but not trained. The model is released under Apache‑2.0.
Good at
- Decision making with typed answers
- Vision inference via included encoder
- Tool‑integrated workflows
Limits
- Requires large VRAM for full precision
- No native vision training support
- Limited to 262144 context tokens
Best for
- Label classification in chatbots
- Automated decision engines
- Vision‑enabled QA systems
Who it’s for. Developers building decision‑oriented applications that need typed outputs and optional vision support. The model is suitable for integration with the shisa‑de client or any OpenAI‑compatible endpoint that can expose logprobs.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 10.6 GB | 19.1 GB | Too big | Get file → | |
| Q3_K_S | 12.2 GB | 20.8 GB | Too big | Get file → | |
| Q3_K_M | 13.3 GB | 21.8 GB | Too big | Get file → | |
| Q3_K_L | 13.8 GB | 22.4 GB | Too big | Get file → | |
| IQ4_XS | 14.1 GB | 22.6 GB | Too big | Get file → | |
| Q4_K_S | 15.5 GB | 24.0 GB | Too big | Get file → | |
| Q4_K_M | 16.8 GB | 25.3 GB | Too big | Get file → | |
| Q5_K_S | 18.0 GB | 26.5 GB | Too big | Get file → | |
| Q5_K_M | 19.1 GB | 27.7 GB | Too big | Get file → | |
| Q6_K | 22.6 GB | 31.2 GB | Too big | Get file → | |
| Q8_0 | 26.9 GB | 35.4 GB | Too big | Get file → |
For 8‑GB VRAM use Q2_K (10.58 GB, 19.1 GB memory at 8k context). For 12‑GB VRAM use Q3_K_S (12.22 GB, 20.8 GB). For 16‑GB VRAM use Q4_K_M (16.8 GB, 25.3 GB).
What GPU do you need for Shisa De 1?
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_S (15.5 GB file, about 24.0 GB in use).
- 32 GB VRAM: the best quant that fits is Q6_K (22.6 GB file, about 31.2 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (26.9 GB file, about 35.4 GB in use).
Which Shisa De 1 quant should you choose?
Q4_K_M is the usual starting point for Shisa De 1: it balances file size and quality well. Here it is a 16.8 GB file.
The Q8_0 file is 26.9 GB, about 1.6 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 10.6 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 Shisa De 1
How much VRAM does Shisa De 1 need?
About 25.3 GB at Q4_K_M with an 8K context window. That is the 16.8 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 Shisa De 1 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 19.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Shisa De 1 GGUF?
It depends on the quant: Q2_K is 10.6 GB, Q3_K_S is 12.2 GB, Q3_K_M is 13.3 GB, Q3_K_L is 13.8 GB, IQ4_XS is 14.1 GB, Q4_K_S is 15.5 GB and 5 more.
What context length does Shisa De 1 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 Shisa De 1 GGUF?
The files are on Hugging Face in the mradermacher/shisa-de-1-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 Shisa De 1 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 Shisa De 1 on 8 GB VRAM?
Use the Q2_K quant. It is 10.58 GB and requires about 19.1 GB memory at 8k context, which is close to the limit of 8 GB VRAM plus CPU fallback.
Does Shisa De 1 support vision tasks?
The vision encoder is present in the weights but was not trained. It can be used for inference but may not perform well on new vision data.