LL

shisa v2.1 llama3.3 70b

LlamaGGUF by mradermacherUpdated 2025-12-09
Download on Hugging Face
Parameters70.6B
Context128K
Layers80
Architecturellama
Smallest quant26.4 GB
Quick answer

shisa v2.1 llama3.3 70b needs about 45.7 GB of memory at Q4_K_M (a 42.5 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 11 quants, from 26.4 GB (Q2_K) to 75.0 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 10.7 GB.

Shisa V2.1 Llama3.3 70B – large Japanese‑English chat model

Shisa V2.1 is a 70.6B parameter Llama3.3 model trained for Japanese and English. It supports a 131072 token context window and is released under the llama3.3 license. The model is available in multiple GGUF quantizations.

Good at

  • Japanese language tasks
  • English chat
  • Instruction following

Limits

  • No vision support
  • Requires large VRAM for high‑quality quants
  • License is llama3.3

Best for

  • Local Japanese‑English chat applications
  • Japanese translation services
  • Edge‑device inference with small quants

Who it’s for. Developers needing a bilingual Japanese‑English LLM for local deployment. The model is suitable for users who can manage the memory requirements of a 70B parameter model and who prefer open‑source licensing.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K26.4 GB 29.6 GB
Too big Get file →
Q3_K_S30.9 GB 34.1 GB
Too big Get file →
Q3_K_M34.3 GB 37.5 GB
Too big Get file →
Q3_K_L37.1 GB 40.3 GB
Too big Get file →
IQ4_XS38.3 GB 41.5 GB
Too big Get file →
Q4_K_S40.3 GB 43.5 GB
Too big Get file →
Q4_K_M42.5 GB 45.7 GB
Too big Get file →
Q5_K_S48.7 GB 51.8 GB
Too big Get file →
Q5_K_M49.9 GB 53.1 GB
Too big Get file →
Q6_K57.9 GB 61.1 GB
Too big Get file →
Q8_075.0 GB 78.2 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (26.38 GB file, ~29.6 GB VRAM). For 16 GB VRAM, Q4_K_S (40.35 GB file, ~43.5 GB VRAM) is a good balance. For 32 GB VRAM, Q5_K_M (49.95 GB file, ~53.1 GB VRAM) offers higher quality.

What GPU do you need for shisa v2.1 llama3.3 70b?

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: no quant fits at an 8K context.
  • 32 GB VRAM: the best quant that fits is Q2_K (26.4 GB file, about 29.6 GB in use).
  • 48 GB VRAM: the best quant that fits is Q4_K_M (42.5 GB file, about 45.7 GB in use).

Which shisa v2.1 llama3.3 70b quant should you choose?

Q4_K_M is the usual starting point for shisa v2.1 llama3.3 70b: it balances file size and quality well. Here it is a 42.5 GB file.

The Q8_0 file is 75.0 GB, about 1.8 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 26.4 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

Architecturellama
Parameters70.6B
Layers80
Trained context131,072 tokens
KV cache (fp16)320 KB per token

Source

Quantized bymradermacher
FamilyLlama
Licensellama3.3
Downloads (HF)1,507
Last updated2025-12-09

Frequently asked questions about shisa v2.1 llama3.3 70b

How much VRAM does shisa v2.1 llama3.3 70b need?

About 45.7 GB at Q4_K_M with an 8K context window. That is the 42.5 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 v2.1 llama3.3 70b on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 29.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of shisa v2.1 llama3.3 70b GGUF?

It depends on the quant: Q2_K is 26.4 GB, Q3_K_S is 30.9 GB, Q3_K_M is 34.3 GB, Q3_K_L is 37.1 GB, IQ4_XS is 38.3 GB, Q4_K_S is 40.3 GB and 5 more.

What context length does shisa v2.1 llama3.3 70b support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 2.7 GB at 8K and 10.7 GB at 32K.

Where can I download shisa v2.1 llama3.3 70b GGUF?

The files are on Hugging Face in the mradermacher/shisa-v2.1-llama3.3-70b-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 v2.1 llama3.3 70b use?

Hugging Face lists the license as llama3.3. 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 V2.1 on 8 GB VRAM?

Use the Q2_K quant; it requires about 29.6 GB VRAM at 8 k context, so 8 GB VRAM is insufficient for inference.

Does this model support vision?

No, Shisa V2.1 does not have vision capabilities.

What license applies to this model?

The model is released under the llama3.3 license.

Other GGUFs of shisa-v2.1-llama3.3-70b

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Llama models

All Llama →

Other large models

All large models →