What are the hardware requirements for Nanbeige 4.3 33B Heretic?

Nanbeige 4.3 33B Heretic requires approximately 65.9 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 Nanbeige 4.3 33B Heretic and how large is it?

Nanbeige 4.3 33B Heretic is developed by Luni, a model with 33b 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 Nanbeige 4.3 33B Heretic?

Nanbeige 4.3 33B Heretic is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Nanbeige 4.3 33B Heretic by Luni — VRAM 65.9GB, 512K context

Name: Nanbeige 4.3 33B Heretic
Author: Luni

Nanbeige 4.3 33B Heretic is an open-source language model by Luni. Features: 33b LLM, VRAM: 65.9GB, Context: 512K, License: apache-2.0, LLM Explorer Score: 0.32.

Anti-slop Base model:finetune:nanbeige/n... Base model:nanbeige/nanbeige4.... Chain-of-thought Conversational De-slopped En Endpoints compatible Full-attention Heretic Llama Nanbeige Reasoning Region:us Safetensors Sharded Tensorflow Thinking Uncensored Zh

Model Card on HF 🤗: https://huggingface.co/Luni/Nanbeige-4.3-33B-Heretic

Nanbeige 4.3 33B Heretic Parameters and Internals

LLM Name	Nanbeige 4.3 33B Heretic
Repository 🤗	https://huggingface.co/Luni/Nanbeige-4.3-33B-Heretic
Base Model(s)	Nanbeige/Nanbeige4.3-33B Nanbeige/Nanbeige4.3-33B
Model Size	33b
Required VRAM	65.9 GB
Updated	2026-10-28
Maintainer	Luni
Model Type	llama
Model Files	4.8 GB: 1-of-14 4.7 GB: 2-of-14 4.7 GB: 3-of-14 4.7 GB: 4-of-14 4.7 GB: 5-of-14 4.7 GB: 6-of-14 4.7 GB: 7-of-14 4.7 GB: 8-of-14 4.7 GB: 9-of-14 4.7 GB: 10-of-14 4.7 GB: 11-of-14 4.7 GB: 12-of-14 4.7 GB: 13-of-14 4.7 GB: 14-of-14
Supported Languages	en zh
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	524288
Model Max Length	524288
Transformers Version	5.20.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<unk>
Vocabulary Size	166912

Rank the Nanbeige 4.3 33B Heretic 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 Nanbeige 4.3 33B Heretic

Best Alternatives	Context / RAM	Downloads	Likes
Nanbeige4.3 33B	512K / 66.3 GB	4785	655
Nanbeige4.3 33B NVFP4	512K / 22.1 GB	537	20
Nanbeige4.3 33B OQ4e	512K / 19.4 GB	894	3
Nanbeige4.3 33B FP8	512K / 4.7 GB	396	2
Nanbeige4.3 33B NVFP4	512K / 4.7 GB	870	0
Nanbeige4.3 33B NVFP4 A16	512K / 22.1 GB	279	1
Nanbeige4.3 33B OQ6e	512K / 27.7 GB	326	1
...anbeige4.3 33B W4A16 AutoRound	512K / 20.8 GB	278	1
Nanbeige4.3 33B EXL3 2.5bpw	512K / 14.4 GB	275	1
Nanbeige4.3 33B JANG 2D	512K / 14 GB	237	1

Note: green Score (e.g. "73.2") means that the model is better than Luni/Nanbeige-4.3-33B-Heretic.