What are the hardware requirements for Hermes 4.3 36B?

Hermes 4.3 36B requires approximately 72.2 GB of VRAM and supports a context window of 512K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Hermes 4.3 36B and how large is it?

Hermes 4.3 36B is developed by NousResearch, a model with 36b 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 4.3 36B?

Hermes 4.3 36B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Hermes 4.3 36B by NousResearch — VRAM 72.2GB, 512K context

Name: Hermes 4.3 36B
Author: NousResearch

Hermes 4.3 36B is an open-source language model by NousResearch. Features: 36b LLM, VRAM: 72.2GB, Context: 512K, License: apache-2.0, LLM Explorer Score: 0.29.

Arxiv:2508.18255 Atropos Base model:bytedance-seed/seed... Base model:finetune:bytedance-... Bytedance seed Chat Chatml Conversational Dataforge Deploy:azure En Endpoints compatible Finetuned Function calling Hybrid-mode Instruct Json mode Long context Reasoning Region:us Roleplaying Safetensors Seed oss Sharded Structured outputs Tensorflow Tool use

Model Card on HF 🤗: https://huggingface.co/NousResearch/Hermes-4.3-36B

Hermes 4.3 36B Parameters and Internals

LLM Name	Hermes 4.3 36B
Repository 🤗	https://huggingface.co/NousResearch/Hermes-4.3-36B
Base Model(s)	Seed OSS 36B Base ByteDance-Seed/Seed-OSS-36B-Base
Model Size	36b
Required VRAM	72.2 GB
Updated	2026-07-29
Maintainer	NousResearch
Model Type	seed_oss
Model Files	5.3 GB: 1-of-14 5.3 GB: 2-of-14 5.3 GB: 3-of-14 5.1 GB: 4-of-14 5.1 GB: 5-of-14 5.1 GB: 6-of-14 5.3 GB: 7-of-14 5.3 GB: 8-of-14 5.1 GB: 9-of-14 5.1 GB: 10-of-14 5.1 GB: 11-of-14 5.3 GB: 12-of-14 5.3 GB: 13-of-14 4.5 GB: 14-of-14
Supported Languages	en
Model Architecture	SeedOssForCausalLM
License	apache-2.0
Context Length	524288
Model Max Length	524288
Transformers Version	4.55.0
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<seed:pad>
Vocabulary Size	155136
Torch Data Type	bfloat16

Rank the Hermes 4.3 36B 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 4.3 36B

Best Alternatives	Context / RAM	Downloads	Likes
Spoomplesmaxx Mockingbird 36B	512K / 72.2 GB	770	0
Seed OSS 36B Instruct	512K / 72.2 GB	45727	505
Seed OSS 36B Base	512K / 72.2 GB	48856	67
Seed OSS 36B Base WoSyn	512K / 72.2 GB	186	53
Seed OSS 36B Instruct	512K / 72.2 GB	382	3
...6B Base Instruct Karcher Merge	512K / 72.1 GB	38	2
Seed OSS 36B Instruct 4bit	512K / 20.2 GB	60	3
Seed Oss 36B Instruct Fp16 Mlx	512K / 72.2 GB	27	2
Seed OSS 36B Instruct AWQ	512K / 21.1 GB	320	8
...eed OSS 36B Instruct GPTQ Int4	512K / 21.1 GB	21	5

Note: green Score (e.g. "73.2") means that the model is better than NousResearch/Hermes-4.3-36B.