LLM EXPLORER 59,420 MODELS INDEXED

Meta Llama 3.1 405B Instruct Bnb 4bit by unsloth

By unsloth · 141 downloads
HF Repo ↗ Copy URL

Meta Llama 3.1 405B Instruct Bnb 4bit is an open-source language model by unsloth. Features: 405b LLM, VRAM: 214.3GB, Context: 128K, License: meta, Quantized, Instruction-Based, LLM Explorer Score: 0.28, ELO: 1335.

  4-bit   4bit   Bitsandbytes   Conversational   En   Endpoints compatible   Facebook   Instruct   Llama   Llama-3   Meta   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Unsloth
Model Card on HF 🤗: https://huggingface.co/unsloth/Meta-Llama-3.1-405B-Instruct-bnb-4bit 

Meta Llama 3.1 405B Instruct Bnb 4bit Benchmarks

LMSys ELO 1335 vs 1272 (so35) 5%
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

Meta Llama 3.1 405B Instruct Bnb 4bit Parameters and Internals

LLM NameMeta Llama 3.1 405B Instruct Bnb 4bit
Repository 🤗https://huggingface.co/unsloth/Meta-Llama-3.1-405B-Instruct-bnb-4bit 
Model Size405b
Required VRAM214.3 GB
Updated2026-07-16
Maintainerunsloth
Model Typellama
Instruction-BasedYes
Model Files  5.0 GB: 1-of-44   4.9 GB: 2-of-44   4.9 GB: 3-of-44   4.9 GB: 4-of-44   4.9 GB: 5-of-44   4.9 GB: 6-of-44   4.9 GB: 7-of-44   4.9 GB: 8-of-44   4.9 GB: 9-of-44   4.9 GB: 10-of-44   4.9 GB: 11-of-44   4.9 GB: 12-of-44   4.9 GB: 13-of-44   4.9 GB: 14-of-44   4.9 GB: 15-of-44   4.9 GB: 16-of-44   4.9 GB: 17-of-44   4.9 GB: 18-of-44   4.9 GB: 19-of-44   4.9 GB: 20-of-44   4.9 GB: 21-of-44   4.9 GB: 22-of-44   4.9 GB: 23-of-44   4.9 GB: 24-of-44   4.9 GB: 25-of-44   4.9 GB: 26-of-44   4.9 GB: 27-of-44   4.9 GB: 28-of-44   4.9 GB: 29-of-44   4.9 GB: 30-of-44   4.9 GB: 31-of-44   4.9 GB: 32-of-44   4.9 GB: 33-of-44   4.9 GB: 34-of-44   4.9 GB: 35-of-44   4.9 GB: 36-of-44   4.9 GB: 37-of-44   4.9 GB: 38-of-44   4.9 GB: 39-of-44   4.9 GB: 40-of-44   4.9 GB: 41-of-44   4.9 GB: 42-of-44   4.2 GB: 43-of-44   4.2 GB: 44-of-44
Supported Languagesen
Quantization Type4bit
Model ArchitectureLlamaForCausalLM
Licensemeta
Context Length131072
Model Max Length131072
Transformers Version4.44.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|finetune_right_pad_id|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to Meta Llama 3.1 405B Instruct Bnb 4bit

Best Alternatives
Context / RAM
Downloads
Likes
...ta Llama 3.1 405B Instruct FP8128K / 197.6 GB55370165
Llama 3.1 405B Instruct FP8128K / 209.2 GB476016
Llama 3.1 405B Instruct128K / 183.1 GB43229595
Meta Llama 3.1 405B Instruct128K / 186 GB55654473
Llama 3.1 405B Instruct FP8128K / 193.4 GB15734193
Shisa V2 Llama3.1 405B128K / 191.2 GB2219
BigLlama 3.1 681B Instruct128K / 189.6 GB1712
INTELLECT 1 Instruct8K / 20.5 GB328125
Note: green Score (e.g. "73.2") means that the model is better than unsloth/Meta-Llama-3.1-405B-Instruct-bnb-4bit.