What are the hardware requirements for Hermes 4 405B FP8?

Hermes 4 405B FP8 requires approximately 213.8 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 Hermes 4 405B FP8 and how large is it?

Hermes 4 405B FP8 is developed by NousResearch, 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 Hermes 4 405B FP8?

Hermes 4 405B FP8 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Hermes 4 405B FP8 by NousResearch — VRAM 213.8GB, 128K context

Name: Hermes 4 405B FP8
Rating: 4 (1 reviews)
Author: NousResearch

Hermes 4 405B FP8 is an open-source language model by NousResearch. Features: 405b LLM, VRAM: 213.8GB, Context: 128K, License: llama3, LLM Explorer Score: 0.21.

Arxiv:2508.18255 Atropos Base model:meta-llama/llama-3.... Base model:quantized:meta-llam... Chat Chatml Compressed-tensors Conversational Dataforge En Endpoints compatible Finetuned Function calling Hybrid-mode Instruct Json mode Llama Llama-3.1 Long context Reasoning Region:us Roleplaying Safetensors Sharded Structured outputs Tensorflow Tool use

Model Card on HF 🤗: https://huggingface.co/NousResearch/Hermes-4-405B-FP8

Hermes 4 405B FP8 Parameters and Internals

LLM Name	Hermes 4 405B FP8
Repository 🤗	https://huggingface.co/NousResearch/Hermes-4-405B-FP8
Base Model(s)	meta-llama/Meta-Llama-3.1-405B meta-llama/Meta-Llama-3.1-405B
Model Size	405b
Required VRAM	213.8 GB
Updated	2026-06-04
Maintainer	NousResearch
Model Type	llama
Model Files	4.8 GB: 1-of-86 4.9 GB: 2-of-86 4.6 GB: 3-of-86 4.9 GB: 4-of-86 4.6 GB: 5-of-86 4.9 GB: 6-of-86 4.6 GB: 7-of-86 4.9 GB: 8-of-86 4.6 GB: 9-of-86 4.9 GB: 10-of-86 4.6 GB: 11-of-86 4.9 GB: 12-of-86 4.6 GB: 13-of-86 4.9 GB: 14-of-86 4.6 GB: 15-of-86 4.9 GB: 16-of-86 4.6 GB: 17-of-86 4.9 GB: 18-of-86 4.6 GB: 19-of-86 4.9 GB: 20-of-86 4.6 GB: 21-of-86 4.9 GB: 22-of-86 4.6 GB: 23-of-86 4.9 GB: 24-of-86 4.6 GB: 25-of-86 4.9 GB: 26-of-86 4.6 GB: 27-of-86 4.9 GB: 28-of-86 4.6 GB: 29-of-86 4.9 GB: 30-of-86 4.6 GB: 31-of-86 4.9 GB: 32-of-86 4.6 GB: 33-of-86 4.9 GB: 34-of-86 4.6 GB: 35-of-86 4.9 GB: 36-of-86 4.6 GB: 37-of-86 4.9 GB: 38-of-86 4.6 GB: 39-of-86 4.9 GB: 40-of-86 4.6 GB: 41-of-86 4.9 GB: 42-of-86 4.6 GB: 43-of-86 4.9 GB: 44-of-86 4.6 GB: 45-of-86
Supported Languages	en
Model Architecture	LlamaForCausalLM
License	llama3
Context Length	131072
Model Max Length	131072
Transformers Version	4.52.4
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Hermes 4 405B FP8 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 Hermes 4 405B FP8

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
Meta Llama 3.1 405B	128K / 186 GB	521433	808
Llama 3.1 405B	128K / 183.1 GB	124477	982
Llama 3.1 405B FP8	128K / 193.4 GB	472107	123
AGI 405B	128K / 191.2 GB	883	3
Hermes 4 405B	128K / 191.2 GB	538	92
Llama 3.1 405B Instruct	128K / 183.1 GB	43229	595
Meta Llama 3.1 405B Instruct	128K / 186 GB	55654	473
Meta Llama 3.1 405B FP8	128K / 197.6 GB	131499	94

Note: green Score (e.g. "73.2") means that the model is better than NousResearch/Hermes-4-405B-FP8.