What are the hardware requirements for Shisa V2 Llama3.1 405B?

Shisa V2 Llama3.1 405B requires approximately 191.2 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Shisa V2 Llama3.1 405B and how large is it?

Shisa V2 Llama3.1 405B is developed by shisa-ai, a model with 405b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Shisa V2 Llama3.1 405B?

Shisa V2 Llama3.1 405B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Shisa V2 Llama3.1 405B by shisa-ai — VRAM 191.2GB, 128K context

Name: Shisa V2 Llama3.1 405B
Author: shisa-ai

Shisa V2 Llama3.1 405B is an open-source language model by shisa-ai. Features: 405b LLM, VRAM: 191.2GB, Context: 128K, License: llama3.1, Instruction-Based, LLM Explorer Score: 0.18.

Arxiv:2501.18511 Base model:finetune:meta-llama... Base model:meta-llama/llama-3.... Conversational Dataset:shisa-ai/deepseekv3-ul... Dataset:shisa-ai/shisa-v2-shar... En Endpoints compatible Instruct Ja Ko Llama Region:us Safetensors Sharded Tensorflow Zh

Model Card on HF 🤗: https://huggingface.co/shisa-ai/shisa-v2-llama3.1-405b

Shisa V2 Llama3.1 405B Parameters and Internals

LLM Name	Shisa V2 Llama3.1 405B
Repository 🤗	https://huggingface.co/shisa-ai/shisa-v2-llama3.1-405b
Model Name	shisa-v2-llama3.1-405b
Base Model(s)	meta-llama/Llama-3.1-405B-Instruct meta-llama/Llama-3.1-405B-Instruct
Model Size	405b
Required VRAM	191.2 GB
Updated	2026-08-08
Maintainer	shisa-ai
Model Type	llama
Instruction-Based	Yes
Model Files	4.8 GB: 1-of-191 4.0 GB: 2-of-191 4.6 GB: 3-of-191 4.6 GB: 4-of-191 3.5 GB: 5-of-191 4.6 GB: 6-of-191 4.6 GB: 7-of-191 3.5 GB: 8-of-191 4.6 GB: 9-of-191 4.6 GB: 10-of-191 3.5 GB: 11-of-191 4.6 GB: 12-of-191 4.6 GB: 13-of-191 3.5 GB: 14-of-191 4.6 GB: 15-of-191 4.6 GB: 16-of-191 3.5 GB: 17-of-191 4.6 GB: 18-of-191 4.6 GB: 19-of-191 3.5 GB: 20-of-191 4.6 GB: 21-of-191 4.6 GB: 22-of-191 3.5 GB: 23-of-191 4.6 GB: 24-of-191 4.6 GB: 25-of-191 3.5 GB: 26-of-191 4.6 GB: 27-of-191 4.6 GB: 28-of-191 3.5 GB: 29-of-191 4.6 GB: 30-of-191 4.6 GB: 31-of-191 3.5 GB: 32-of-191 4.6 GB: 33-of-191 4.6 GB: 34-of-191 3.5 GB: 35-of-191 4.6 GB: 36-of-191 4.6 GB: 37-of-191 3.5 GB: 38-of-191 4.6 GB: 39-of-191 4.6 GB: 40-of-191 3.5 GB: 41-of-191 4.6 GB: 42-of-191 4.6 GB: 43-of-191 3.5 GB: 44-of-191 4.6 GB: 45-of-191
Supported Languages	ja en ko zh
Model Architecture	LlamaForCausalLM
License	llama3.1
Context Length	131072
Model Max Length	131072
Transformers Version	4.50.0
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Shisa V2 Llama3.1 405B capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Shisa V2 Llama3.1 405B

Best Alternatives	Context / RAM	Downloads	Likes
...ta Llama 3.1 405B Instruct FP8	128K / 197.6 GB	55370	165
Llama 3.1 405B Instruct FP8	128K / 209.2 GB	4760	16
Llama 3.1 405B Instruct	128K / 183.1 GB	43229	595
Meta Llama 3.1 405B Instruct	128K / 186 GB	55654	473
Llama 3.1 405B Instruct FP8	128K / 193.4 GB	15734	193
BigLlama 3.1 681B Instruct	128K / 189.6 GB	17	12
INTELLECT 1 Instruct	8K / 20.5 GB	328	125
...ama 3.1 405B Instruct Bnb 4bit	128K / 214.3 GB	141	9

Note: green Score (e.g. "73.2") means that the model is better than shisa-ai/shisa-v2-llama3.1-405b.