POCKET 26B 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 2 quants, from 11.1 GB (Q2_K) to 16.8 GB (Q4_K_M). Longer context needs more: at 32K tokens the KV cache alone adds about 32.2 GB.
POCKET 26B – a Korean‑tuned Gemma4 26B model for on‑device use
POCKET 26B is a re‑quantized version of google/gemma‑4‑26B‑A4B‑it. It retains the full 25.2B parameters and 4B active MoE. The model supports a 262144 token context window and is licensed under Apache‑2.0.
Good at
- Large context reasoning
- Multilingual text generation
- Efficient on‑device inference
Limits
- No vision or audio support
- Requires 8k context for memory estimates
- No GPU‑only runtime needed
Best for
- Local text generation on laptops
- Edge devices with limited VRAM
- Developers needing a Korean‑tuned Gemma4
Who it’s for. Developers and researchers who want a fully on‑device Gemma4 model with Korean tuning. The model runs in mainstream runtimes such as Ollama, LM Studio, PocketPal, koboldcpp, and the browser without requiring a bleeding‑edge build.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 11.1 GB | 19.7 GB | Too big | Get file → | |
| Q4_K_M | 16.8 GB | 25.3 GB | Too big | Get file → |
For 8 GB VRAM, use Q2_K (file size 11.12 GB, estimated 19.7 GB memory at 8k context). If more VRAM is available, Q4_K_M (16.8 GB, 25.3 GB memory) offers higher precision. Memory estimates are approximate.
What GPU do you need for POCKET 26B?
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 Q2_K (11.1 GB file, about 19.7 GB in use).
- 32 GB VRAM: the best quant that fits is Q4_K_M (16.8 GB file, about 25.3 GB in use).
Which POCKET 26B quant should you choose?
Q4_K_M is the usual starting point for POCKET 26B: it balances file size and quality well. Here it is a 16.8 GB file.
If memory is tight, the smallest file is Q2_K at 11.1 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 POCKET 26B
How much VRAM does POCKET 26B 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 POCKET 26B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 19.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of POCKET 26B GGUF?
It depends on the quant: Q2_K is 11.1 GB, Q4_K_M is 16.8 GB.
What context length does POCKET 26B 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 POCKET 26B GGUF?
The files are on Hugging Face in the FINAL-Bench/POCKET-26B-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 POCKET 26B 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 POCKET 26B on 8 GB VRAM?
Yes, the Q2_K quantization requires about 19.7 GB memory at 8k context, which fits on 8 GB VRAM with swap or on a CPU.
Does POCKET 26B support image input?
No, this model does not include vision or audio capabilities.
What license applies to POCKET 26B?
The model is released under the Apache‑2.0 license.